SAN MATEO (07/03/2000) - It always seems to come down to the same issue regardless of how often technology models change. Take for instance the e-business phenomenon, which has helped spawn new software categories such as digital exchanges, corporate portals, and e-commerce applications. But when you start to peel back the onion, you find yourself in a familiar discussion that goes like this: Should your organization go with a suite of products from one vendor that promises tight integration, or should it pursue a best-of-breed strategy relying on emerging Web standards to tie it all together?
For those of you who missed the early part of the last decade, this is the same discussion that led to the dominance of Microsoft Corp. Office over rival applications from WordPerfect and Lotus Development Corp. on the desktop and the dominance of SAP AG in the ERP (enterprise resource planning) category.
A lot has changed over the last two years with the emergence of de facto Java and XML standards. But many questions remain as to just how far these standards can drive a best-of-breed approach to applications.
For example, Java applications rely heavily on application servers. But right now there are more than 40 application servers available, and every one of them handles persistence differently, which makes integration a challenge.
Meanwhile, the companies driving XML in the digital exchange space, such as Ariba and CommerceOne Inc., have a habit of defining their own XML and DTD (Document Type Definition) implementations. This makes integrating multiple exchanges a challenge. In fact, multiple companies have released exchange integration tools, yet another form of middleware.
Alas, neither one of these situations looks as though it will get better any time soon. It's highly unlikely that Sun has the political weight to get other vendors to agree to a common persistence model because of the angst among vendors dealing with Sun's approach to Java licensing fees. At the same time, the World Wide Web Consortium(W3C), charged with the stewardship of XML, is now exhibiting the trappings of a large-scale bureaucratic industry movement (it now takes this organization three times as long to accomplish something as it did a year ago).
This, of course, is the exact state of affairs that companies such as Oracle, IBM, and SAP are hoping to play into. Each company is being out-hustled by rival start-ups in every aspect of the e-business software market, so to compensate they have come up with variants of integrated suites that promise to make the customer's integration headaches go away.
For example, Oracle argues that its newly launched portal offering fits perfectly with the rest of its wares. SAP meanwhile positions mySAP.com as a natural extension to its client/server-based suite of ERP applications, and IBM would have us all believe that it alone can integrate your existing legacy applications with a new generation of Web-based apps developed on top of WebSphere.
The spin factor put forth by these companies is pretty high, but there is some merit to each of their arguments. The challenge for IT people is to decide what tradeoffs are most appropriate for their organizations.
For example, there's no question that your entire IT organization, from the front-office group using CRM (customer relationship management) to the supply-chain management systems in manufacturing, needs to be integrated with a wide range of systems that exist both inside and outside your company. But the approach to that problem is going to differ widely.
One company may decide that a best-of-breed corporate portal can be sacrificed in the name of keeping the number of vendors involved to a minimum. Another company may decide that it has the internal expertise needed to sustain a best-of-breed approach -- a strategy that is a lot easier today than it was two years ago. There are no hard and fast rules here, except to use common sense.
So do your homework, turn off the TV ads, and think long and hard about what kind of IT organization you are in. Once you decide that, your course of action becomes relatively clear.
Michael Vizard is editor in chief of InfoWorld.