Federal XML group updates guidelines
- By Joab Jackson
- Sep 23, 2005
'We want to share that knowledge so agencies won't have to reinvent it.'
'Owen Ambur, XML Community of Practice
Three years is a long time in the field of information management. So it's not surprising that the Federal XML Community of Practice, along with the General Services Administration and the Homeland Security Department, is revamping a widely used federal guide for creating Extensible Markup Language-based schemas.
Since the 2002 release of the XML Developers Guide, there has been a lot of 'additional thought and intellectual investment made into data architecture, by the Navy and other agencies,' said Owen Ambur, co-chair of the CIO Council's XML Community of Practice and chief XML strategist for the Interior Department. 'We want to share that knowledge so agencies won't have to reinvent it.'
Not all the changes come with unequivocal support by users of the current guidelines, however. The Intelligence Community Metadata Working Group submitted a paper to the group detailing its issues with the new draft. Chaired by Tim West of the Defense Intelligence Agency, the working group represents 15 U.S. federal intelligence agencies. The group now estimates that the changes in the guidelines would affect up to 35 projects within its agencies.
'We do not believe this document presently meets the needs of the [intelligence community]. The proposed rules are too restrictive for the many applications of XML currently in use or under development,' the group stated in the paper.
GSA's Office of Governmentwide Policy is overseeing the effort, along with the DHS' Metadata Center of Excellence. On Oct. 19, the working group will submit a draft of the revised document to the federal XML group. If approved, it will be handed up to the council's Architecture and Infrastructure Committee. The committee may submit the guidelines to the Office of Management and Budget as proposed policy or general guidance.
GSA oversaw the first draft of the guidelines, which was primarily written for developers creating XML schemas for agency programs. The idea behind the document was to encourage agencies to create schemas that could work well together and ease exchange of data. The document handles low-level details such as the syntactic construction of element names. Such detail 'needs to be addressed to achieve machine-to-machine interoperability,' Ambur said.
The group is taking a number of existing guidelines as the basis for the re- vamped document, according to Mark Crawford, a senior research fellow for LMI Consulting of McLean, Va. LMI supports GSA in XML projects.
The updated guidelines'to be renamed Federal XML Naming and Design Rules and Guidelines'emphasize a modular approach to developing federal schemas, Crawford said (See box). In addition to just supplying guidelines, the revision also will provide examples agencies can use to craft their own schemas.
The guide will borrow from outside sources. Like its predecessor, it will borrow heavily from individual agency guidelines, such as those authored by the Navy, the Environmental Protection Agency, the Internal Revenue Service and Justice Department.Standards consulted
It also will incorporate material from standards bodies, most notably the Naming and Design Rules guidebook and the Universal Business Language specifications from the Organization for the Advancement of Structured Information Standards. Material from the U.N. Centre for Trade Facilitation and Electronic Business also will be included.
The Intelligence Community Metadata Working Group has a variety of concerns about the revision, from the very general to the specific. The group expressed concern that the guidelines borrow too heavily from business schemas like UBL and the electronic Business Extensible Markup Language, both of which are primarily used for conducting electronic business transactions. Such rules are not '100 percent applicable to ... all XML applications of all federal agencies,' the paper states.
The IC working group also cited a number of specific concerns, such as the banning of acronyms as schema elements. The group noted that in many cases, an acronym of a program is not classified, while the full title is.
'We've taken a look at those concerns and tried to alleviate them to the maximum extent possible,' Crawford said.