Tom Welsh Senior Consultant

Tom Welsh is a Senior Consultant with Cutter Consortium's Enterprise Architecture advisory service and former Editor of Cutter Consortium's monthly Web Services Strategies. He is an independent consultant and analyst specializing in middleware, object technology, and software engineering. At Digital Equipment Corporation, Mr. Welsh was a hardware technician, software support specialist, corporate software developer, and senior technology consultant, before taking on the task of marketing Digital's OO software products in the UK. In his role as a principal analyst at ComputerWire, Mr. Welsh has been a leading contributor to Client/Server Tools Bulletin, Internet Tools Bulletin, and Object-Oriented Tools Bulletin. He has written many reports and papers as well as chaired and spoken at conferences and seminars. Since 1992, Mr. Welsh has closely followed the work of the OMG and its specifications, including CORBA, UML, XMI, and CWM. He can be reached at consulting@cutter.com.

XML: One Metalanguage to Rule Them All?

An interview with Tom Welsh, Senior Consultant, Cutter Consortium

In this interview, Tom Welsh addresses the misconceptions, impacts and future potential of Extensible Markup Language (XML).

Q: What common misconceptions do people have about XML?

There is a small, hard-core community of architects and developers who do a lot of work with XML. Those people understand it intimately, and know its limitations. Pretty well everyone else is laboring under a variety of mistaken beliefs about it. My personal favorite is the Wall Street VP who told a journalist: "We are going with XML because it is the fastest Internet protocol!" Many people in the media, who are not developers themselves, hear XML being called a "language" and jump to the conclusion that it is a programming language like C or Java.

Then there are the subtler misunderstandings, like the often-heard notion that data in XML format is "self-describing." To a human being, perhaps; to a computer program, absolutely not. That, in turn, leads on to the talk about "loose coupling," which tends to be wildly exaggerated.

But the two biggest misconceptions are also the simplest and least technical: that XML is new and revolutionary, and that it simplifies the business of creating and using software. "Hope springs eternal in the human breast," but the facts are that XML is a new edition of 1970s technology; and that, far from simplifying anything else, XML itself is already becoming dangerously complicated. Decision-makers who do not actually use XML like to think that it will somehow make their lives easier, so they tend to see it as a hammer that will change all their problems into nails.

Q: Tell us about some of the current efforts to promote XML.

One of the most fascinating things about XML is the way it has managed to unite everyone behind it. You have to go all the way back to the way relational databases swept the market in the early 1980s to see anything similar, but even they faced heavy opposition early on. In addition to its technical merits, XML has probably benefited from being created at the World Wide Web Consortium (W3C), and then taken up by Microsoft and, shortly after, IBM. That is the kind of quorum that vendors just cannot afford to resist.

Everyone at W3C has been pushing XML, of course, even in long-range projects like the Semantic Web. Then, Microsoft emphasized it as the key aspect of .NET, with much talk about the importance of XML Web services -- a campaign that has been joined by rivals like BEA, Computer Associates, HP, IBM, Oracle, SAP, Sun. Sun, in particular, made sure Web services were built into J2EE 1.4, and the big application vendors like PeopleSoft, SAP and Siebel have been engineering XML interfaces into their new product lines. New XML database products are springing up, and all the leading relational database management systems are getting XML capability added on.

Then there is a horde of specialized initiatives, from Electronic Business XML (ebXML) and Interactive Financial eXchange (IFX) to eXtensible Business Reporting Language (XBRL) and exotic initiatives like VoiceXML.

Q: How might those efforts impact the application and use of XML in the short term?

Any innovation as complex and popular as XML behaves like a big wave. It breaks in a welter of foam, then runs way, way up the beach. Then it ebbs way back, as people begin to worry that they may have expected too much. Right now, XML is permeating just about everything in the software world. There is some risk of over-exuberance, chaos, and loss of control. But that is the whole charm of free-enterprise capitalism! Lots of things get tried -- some work, others don't. I think the key to reacting appropriately lies in understanding XML, rather than giving way to uninformed enthusiasm. In a way, it's just like investing in the stock market: Gurus like Warren Buffett will tell you never to put your money into a business that you, personally, do not thoroughly understand.

Q: Where do you see XML being positioned five years from now?

There can be no doubt that XML will permeate the software industry -- not just because of its technical strengths, but also through the unprecedented cooperation it has generated. Software development, data management, operations and system management, applications... it will be everywhere. It has even been suggested that e-mail might be XML-based in future, with great potential benefits. Some of the more optimistic predictions about Web services, for instance, may not be fulfilled -- not for many years, at least. Much depends on the ongoing efforts to define shared vocabularies, or ontologies; that is a really fundamental challenge.

In due course, something else will come along and take over the limelight, and XML will fade into the background as just another tool in the software toolbox. Only then will we really be able to assess its true value.