Think you have a content management problem? The federal government has 33 million Web pages scattered across 23,000 Web sites--a mountain of information choking in redundancy, antiquated systems, and muddled information architectures.
Sworn in last fall, the first CTO of the Office of Management and Budget (OMB), Norman Lorentz, has pledged to clean up this gargantuan mess. And Mark Forman, the OMB's associate director for information technology and e-government, says the fed plans to start offering secure Web access to such confidential information as social security benefits later this year.
Such progressive noises sound promising, but the General Accounting Office (GAO) has clearly adopted a go slow policy in one vital technology area: XML. That's too bad, because--in my dreams, anyway--I think the federal government could play a valuable role in Web services.
How? By publishing government documents valuable to business and the general public in XML format in a UDDI directory. Sound far-fetched? Well, not as far-fetched as rearchitecting 23,000 Web sites--just ask anyone who has tried to navigate the politics of fixing just one big corporate Web site. Lorentz's Herculean housecleaning simply won't be completed in our lifetime. This may sound crazy, but I think he has a much better chance of setting de facto XML standards where industry consortiums have yet to hash things out.
Consider, for example, all the data issued by the U.S. Census Bureau or the heap of daily, weekly, and monthly material published by the Federal Reserve Board. Most of that stuff is now published in PDF files. Imagine the benefit if marketers could consume demographics or corporate economists could devour macroeconomic data as Web services instead.
The GAO made a blanket statement that XML vocabularies haven't been hashed out sufficiently, tossing XML into the government's "further study" bin. But that assertion has already been challenged by at least one organization--the HR-XML Consortium, which complains that the GAO has ignored its efforts in developing XML vocabularies for the HR industry. Although there are 24 e-government initiatives in the works, at this point, it appears that all we can expect from them on the XML front is a handy government directory of the existing XML vocabularies themselves.
What a wasted opportunity. Today, most of what we hear about Web services and XML involves either integration inside the firewall or point-to-point data exchange between businesses. No one has yet figured out a viable business model for actually supplying data via Web services, as demonstrated by the existing UDDI directories of largely toylike Web services (such as those maintained by, say, IBM or the two stalwart independents, XMethods and Salcentral).
A concerted, accelerated government XML initiative could help in three ways. First, the effort would expose information--for which we've already paid in tax dollars--in a format usable by business. Second, working with industry consortiums, the government could help establish de facto XML vocabularies simply by publishing useful documents using agreed-upon sets of tags--after all, hashing out these vocabularies is the biggest stumbling block to machine-to-machine communications. Finally, even though businesses can derive huge benefits from consuming the data, it's tough to see how Web services data suppliers will be profitable. That's the cue for Uncle Sam to step in.
Should the government give Web services a leg up? Or should it keep its hand off? E-mail Eric or TalkBack below.
