X
Business

Building a global B2B standard

UDDI is a concept directory currently being developed by Ariba, IBM and Microsoft. If successful, it may prove to be an indispensible tool for doing world-wide business.
Written by Terence Goggin, Contributor

UDDI is a concept directory currantly being developed by Ariba, IBM and Microsoft. If successful, it may prove to be an indispensible tool for doing world-wide business.

At its most far-reaching, UDDI is a means of allowing e-commerce-based businesses to automatically interact and select partners or affiliates.

Imagine a single online phone book that lists product and contact information for every company in the world. Few companies would dare to build the foundation for such a system, but Ariba, IBM and Microsoft are going to give it a try. The trio is hard at work on a free e-commerce service known as UDDI (Universal Description, Discovery and Integration).

According to a white paper released by the companies, UDDI is "a global, platform-independent, open framework to enable businesses to: (1) discover each other, (2) define how they interact over the Internet, and (3) share information in a global registry that will more rapidly accelerate the global adoption of B2B e-commerce."

Or, in somewhat more understandable terms, UDDI will serve as a comprehensive, global directory of businesses organized by service or product. At its most far-reaching, UDDI is a means of allowing e-commerce-based businesses to automatically interact and select partners or affiliates.

Let's take a look at the UDDI directory and its related services to see how close it comes to meeting its goals.

Still confused?

Think of UDDI first and foremost as a giant registry, or directory, of companies and services - not unlike a search engine or a phone book.

Clearly, a large, worldwide directory of this type would be of immense value to businesses everywhere.

But, of course, there are hundreds of such directories and search engines in existence already.

What makes UDDI different in this regard is that it's actually three registries, each of which can be related back to the other two. (That differentiates it from most search engines, which are based strictly on keyword searches.)

The first of these registries, known as the "White Pages," is a simple, straightforward bulk listing of all the businesses known to the system. A business entry in the White Pages might consist of a company's name, primary street address, contact information, and other standard business identifiers such as D&B numbers, tax identifiers, and so on.

The second registry, the "Yellow Pages," is a collection of business products and services, which are associated with a particular business. Here, a business can register all of its services or products.

The third and final registry, called the "Green Pages," is a more technically oriented registry in which a business can describe how another business should interact with it.

Of those registries, the most exciting to businesspeople probably will be the Yellow Pages, because it uses relevant business classifications.

In the past, most businesses in the United States had been categorized according to service or product class based on something called an SIC (Standard Industrial Classification) code.

SIC codes, however, were last updated in 1987, and are woefully inadequate at representing today's technology-oriented businesses. (According to at least one database organized by SIC codes, Yahoo was listed as an ISP.)

UDDI is important and interesting because, if for no other reason, it will classify businesses according to the newer, more relevant, and more frequently updated NAICS (North American Industry Classification System) codes.

Clearly, a large, worldwide directory of this type would be of immense value to businesses everywhere.

More than a portal

Aside from its registries and their organization, UDDI's real importance comes from the services it provides to allow access to those registries. UDDI isn't just a set of registries; it's also an entire API (applications programming interface) that allows users, B2B exchanges or traditional applications to search the registries, presumably for the purpose of finding compatible business partners.

The entire architecture of this API is based on HTTP, XML, Simple Object Access Protocol (SOAP) and other emerging cross-platform data standards.

 

This is relevant because other businesses (and potential customers, partners, etc.) now will be able to find your business more easily. Further, because the information describing your business is more accurate than it probably was in previous directories or search engines, the type of inquiries and business you'll get should be much more targeted to what it is that you do.

The entire architecture of this API is based on HTTP, XML, Simple Object Access Protocol (SOAP) and other emerging cross-platform data standards.

That means UDDI will be just as relevant whether your e-commerce systems are designed to work on Windows, Linux, mainframes running OS/390 - or anything else for that matter.

Incidentally, HTTP, XML and SOAP are some of the same core technologies being used to build the next generation of development tools from Microsoft, known as .NET.

What's the catch?

UDDI sounds promising, but it's still in a development phase, and there are a number of issues to be worked out. In fact, one of the white papers from its three developers even contains an apparently accidental comment by someone working on UDDI's design, which says, "Somebody's gotta settle this one soon!"

Now would be a good time to start investigating how you can benefit from UDDI

 

Additionally, there is always the testing phase, which inevitably will uncover some minor problems that need solving.

So, while there is a specification that describes how UDDI will operate, there's still a little work yet to be done. Although the UDDI concept was developed and released by Ariba, IBM and Microsoft, the eventual intent is to turn the specification and on going service definition to a vendor-neutral, third-party organization.

This entity will manage the intellectual-property issues and further the specification. The initial specification is out, and plans for at least two subsequent revisions have been slotted. Beta implementations that interoperate with each other are expected to be available sometime this month.

We suspect that you will begin to see the first examples of the registry appear within the next few weeks.

After that, expect to see revisions that add functionality and flexibility to the directory over the next 12 to 15 months. At that point, control probably will be transferred to a third-party standards organization.

But there's no reason to wait until the cake is out of the oven. Now would be a good time to start investigating how you can benefit from UDDI - either by helping your partners get their classifications in order and posted or by learning how to integrate your customers' business systems with the registry.

The UDDI specifications are available for anyone to download from the Internet for free, and open-source implementations already have begun. One of these efforts is the jUDDI Project.

Certainly the fact that it's free to participate means it's a low-risk experiment - with lots of potential upside.

Editorial standards