XML marks the spot
- By Megan Lisagor
- Apr 15, 2002
E-government and Extensible Markup Language appear to have crossed paths at a most convenient time.
Better known as XML, the tool for exchanging documents and data has emerged in the past 18 months as something of a universal language for the Internet — as long as everyone agrees on the working vocabulary.
If agencies ever needed a universal language, the time is now. Even as XML has gained momentum during the past year, the Office of Management and Budget has been formulating an e-government strategy that hinges on cross-agency transactions.
"XML would be a key solution of any segment of e-government we go with," said Mayi Canales, deputy chief information officer at the Treasury Department and an e-government portfolio coordinator at the CIO Council.
OMB's 24 e-government initiatives aim to improve service to citizens by eliminating redundant systems and creating one-stop shops for everything from disaster assistance to tax filing. XML provides a universal format for data exchange and, therefore, is the tool needed to get the job done, many say.
"I think you're going to find a little bit of XML in all of the initiatives," said Lew Sanford, e-government program manager for the General Services Administration. "This is what it was designed to do."
The federal government, however, has only begun tapping the technology's potential, officials said.
"It's being presented as an opportunity," he said. "It's going to get stronger emphasis on certain kinds of projects. If everybody was starting from scratch today, they would probably all run out and find their XML approach."
Many agencies, though, are not starting from scratch. E-government and XML clearly are converging, but XML remains a complex solution, requiring careful planning — something that in most instances has not happened yet in the federal government.
A Strong Case
Some observers say that although the e-government initiatives may not have been picked with XML in mind, they might as well have been.
"The first time I saw the list of proposed projects, my remark was, 'It's a set of XML schemas'" or documents, said Owen Ambur, co-chairman of the XML Working Group established by the CIO Council.
At the very least, XML experts say, there's a strong case for agencies to adopt the technology if they are serious about e-government.
"XML is absolutely critical to enabling e-government. It will not happen without XML," said Michael Daconta, co-author of "XML Development with Java 2." "The government especially needs to be more aggressive."
XML, after all, isn't new. It has been a World Wide Web Consortium (W3C) standard since 1998 and is based on the Standard Generalized Markup Language (SGML) developed in the early 1980s and adopted widely in government.
The designers of XML built on SGML's strengths to create a meta language that describes information by giving programmers a common way to tag or mark up documents. An application that receives an XML file automatically knows how the data is structured, making it easier to process the information and shuttle it between systems.
"People think it's a language in itself," said Daconta, who is also director of Web and technology solutions at McDonald Bradley Inc. But in reality, "it's a toolkit for developing languages."
For instance, there are LegalXML, QueryXML and VoiceXML, among others. One standards group has even begun developing a version of XML to manage biometric information (see "XML biometric standards jell," Page 19).
"The fact that it's open and interoperable means we can exchange information that any program now or in the future can read," said Liam Quin, XML activity lead at W3C.
'Mix and Match'
"We are still learning, experimenting," said Marion Royal, co-chairman of the XML Working Group. "XML is probably in its early adulthood."
The same could be said about the 24 e-government initiatives, meaning not all of the agencies have determined what technology they will use to reach their goals.
"If we're doing development, we would expect to use XML," Sanford said. "If it's old, we're assessing whether we should."
Some places are an obvious fit. "Wherever you have systems exchanging data, you have an opportunity to look at XML as the enabler for architecture," Royal said.
For example, at least three of GSA's projects — federal asset sales, integrated acquisition and e-authentication — will benefit from the markup language. "It's a mix and match," Sanford said. "We're trying to do a very leveraged strategy."
Likely candidates include Treasury's EZ Tax Filing, the National Archives and Records Administration's e-Records Management and the Office of Personnel Management's Enterprise Human Resources Integration.
XML also is part of GSA's plans to incorporate content management capabilities into the newly redesigned FirstGov Web site, which will house many of the initiatives, according to Ambur.
Still, a recent General Accounting Office study of XML takes GSA and the National Institute of Standards and Technology "to task for not being more directive," he said.
The report said that despite multiple initiatives to define common federal standards and requirements for XML, the lack of central XML guidelines could derail interoperability within government.
Every user community must create business standards that provide the vocabulary to perform transactions, and federal agencies have not come together to define the governmentwide vocabulary, according to the GAO report, released April 5.
Without this central effort or any attempt to encourage widespread definitions, "the use of XML in the federal government may have only limited benefits and may not achieve the technology's promise of facilitating broad interoperability among disparate systems," the report stated.
Several agencies oversee federal information policy and standards, starting with OMB but also including NIST, GSA and the Defense Information Systems Agency.
Each is involved in developing some form of government XML business standards, and that approach will not help foster the intended governmentwide interoperability, according to the report.
GAO is recommending that OMB, along with NIST and the CIO Council's XML working group, develop a strategy for governmentwide adoption of XML to guide agencies and ensure that XML is incorporated into each agency's enterprise architecture.
"The biggest problem with the federal government is [that] it has no profit incentive to come to consensus," Daconta said. "This actually [requires] congressional mandates." Without a mandate, agencies will be slow to adopt XML, he said.
But officials say all agencies are at least heading in the direction of XML. "There are pockets of prototyping, but it's the wave of the future," Canales said. "Everybody is experimenting with it."
"What should we do about this XML?" Royal recalled NASA CIO Lee Holcomb asking the CIO Council's Architecture and Infrastructure Committee.
The members' response was to form a working group in June 2001 to facilitate the efficient and effective use of XML governmentwide. Responsibilities include education, outreach, forging partnerships and, most important, identifying best practices.
"It's fine to have all this XML running around all over the place, but somewhere you need to be able to find it," said Lynne Rosenthal, manager of the standards and conformance testing group at NIST.
With NIST's support, the XML Working Group is developing a registry of "inherently governmental" data elements, document type definitions and schemas.
"It's a de facto priority, based upon funding," Ambur said. "Hopefully, it will help agencies work more effectively together."
The idea is to create a place where federal workers can register their concepts and see what others are doing with XML.
"If everyone creates different tag names, it makes interoperability very difficult," Royal said.
A council study last year found too many vocabularies and no clear leader in terms of a good set to use, he said. Now the group is working with the Navy, which has a long history with SGML and more recent experience with XML, to provide recommendations to developers at agencies.
"Absolutely, there does need to be an agreed-upon set of syntax, and it needs to be shared," said Patrick Gannon, president and chief executive officer of the Organization for the Advancement of Structured Information Standards (OASIS), a nonprofit international consortium.
OASIS — the home of e-business standards — operates its own clearinghouse at www.xml. org. The smaller-scale, federal version is one way to keep agency information open, Gannon said.
The federal registry is "not intended to be policy. We just encourage people to look at it," said Lisa Carnahan, a computer scientist at NIST, adding that she would like to see a private vendor run the project.
The working group favors using industry standards whenever possible. "There's no reason for us to reinvent the wheel," Ambur said.
According to Michael Daconta, co-author of "XML Development with Java 2," Extensible Markup Language makes sense for e-government because:
* It enables applications to be interoperable.
* Data becomes self-describing.
* Information can be separated from the way it is presented.
* Information can be presented to a variety of end users regardless of the device they're using to view it.
Getting the services "to the citizens is all about integrating smartly," said Daconta, who is also director of Web and technology solutions at McDonald Bradley Inc.