XML Working Group

Meeting Notes

April 21, 2004

The meeting was hosted by Microsoft its office in Bethesda, Maryland.

Owen Ambur announced his intention to enhance the Department of the Interior’s (DOI) glossary of technical terms taking into account ISO 11179 and perhaps using XML to leverage similar glossaries on the Web. http://www.doi.gov/ocio/architecture/appendixa.htm He will also be drafting an XML policy for DOI patterned after the policy issued by the Navy in December 2002. http://xml.gov/documents/completed/donXMLpolicy.pdf

Ken Sall asked about the status of updates to the XML Developers Guide issued in draft form by the Working Group in April 2002. http://xml.gov/documents/in_progress/developersguide.pdf Owen indicated that requests have been made but that resources have not been forthcoming to update and formally issue the Guide through the CIO Council.

Owen announced that the Interagency Committee on Government Information authorized by the eGov Act is seeking comments, by April 30, on its draft Web content policies. http://www.cio.gov/documents/ICGI.html He also noted that Betty Harvey of the DC Area XML User Group and he have initiated plans for a fall forum on XML content management applications similar to the XML authoring/editing tools forum conducted at the Key Bridge Marriot on September 29, 2003. http://www.eccnet.com/xmlug/XML-forum/vendor-info.html

Finally, depending upon agenda priorities and meeting space, Owen noted that he is contemplating combining the May meeting of the Working Group with Sol Safran’s IRS XML stakeholders meeting on May 18.

Lisa Ruff of Microsoft briefed the group and demonstrated some of the XML capabilities of InfoPath and the Microsoft Office Professional Suite. She talked about the need for “smart client” software to unlock proprietary data for connection to other applications and processes. She noted that InfoPath has been included in the Suite under agreements with some of Microsoft’s large customers, including government agencies. She pointed out that XML simplifies data sharing and Microsoft is bringing XML to the desktop. Any XML schema (XSD) can be brought into Office Professional applications and mapped to the structure of the user’s files. XML templates can be saved in Word and WordML provides for “round tripping” with respect to preservation of formatting. FrontPage has been tightly integrated with SharePoint for collaboration.

By using templates users can share “smart” (readily shareable) data without having to know anything about XML. It is easy to bring data into files via a Web Service or a simple URL. Templates can be updated once and all of the related files can be automatically updated. Validation of data occurs upon saving the file. Reports can be generated automatically from data or content repositories. MSDN provides 123 pages of helpful information and 15 lab exercises on building and using InfoPath forms and applications. InfoPath and the Office Suite enable users to work offline while at the same time being able to readily transmit XML formatted data to online applications. XML schema can be imported or created from an existing form. Digital signatures can be applied to forms or to individual sections of forms in InfoPath. Handwriting-to-text conversion is also supported.

Lisa demonstrated how easily XML tagging can be accomplished in Word as well as how easily the presentation can be controlled and changed by applying XSLT. Using the Windows XP task pane, the elements of schema can be mapped to an existing file and then saved as an XML template, simply by highlighting and clicking. She did not have time to demonstrate Excel but noted that data element mapping is done in the same manner as in Word. It is very easy to import and export XML, and lists can be sorted within an Excel spreadsheet. In Word the XML tags can be toggled on or off. In the tags-on mode, the XML tags are displayed in similar fashion to WordPerfect’s reveal codes mode. Word also provides the capability to lock down sections to prevent them from being changed. The save-as-XML option is appropriate for sharing the data, such as when submitting data for inclusion in a database, whereas the WordML rendition is appropriate for printing.

Lisa said InfoPath plays very well with Sharepoint portal server. She briefly displayed a leave request form from the InfoPath forms library. InfoPath looks and works like other Office Suite applications. The task pane can contain context sensitive instructions. Calendar controls can be dragged and dropped on date fields. Forms can be spelled checked and, when additional lines or fields are needed, they can easily be inserted by the user. With BizTalk and InfoPath, workflow can be automated and it can be made available via the Web by using SharePoint.

Lisa’s presentation is available at http://xml.gov/presentations/ms/mosXML_files/frame.htm

Sol Safran briefed the group on IRS’s XML-related activities. He noted that CSC is the prime integrator, assisted by IBM, Northrop Grumman, and others. He said XML is mushrooming and that it is being used both for the receipt and storage of tax returns, which must be maintained in the form in which they are received. Companies want to file electronically and, as the volume grows, the verbosity of XML will become an issue. However, XML provides for improved interoperability, data quality, and taxpayer interaction. Among the tools required is an XML registry.

Ken Sall asked if the XML standards that IRS has adopted are publicly available, and Sol indicated he will check to see if they can be made available through the XML Working Group. He noted the need for namespace management standards, XML component standards and guidelines, as well as the need to standardize IRS global data types. In addition, he suggested the need for a policy on what to XML-ize and what not to. He reiterated that XML is huge (verbose) and getting bigger, and that it will be necessary to be selective in choosing which data to render in XML – with reference to performance issues. He also reiterated that IRS must have an XML registry, since it has big schemas and many of them.

Ken asked if IRS is considering using CORE.gov. Sol said it will be considered along with the DoD registry and others. Owen noted that CORE.gov is not an XML registry per se nor is it compliant with ISO 11179. He noted that GSA funded the business case for the XML registry and that $2.1 million had been included in the President’s budget request. When he was at OMB, Mark Forman had invited other agencies to join GSA in cosponsoring the business case. Bob Benedict commented that NASA has been prototyping the DoD registry and is learning from it. Owen expressed frustration that OMB has fallen short in exerting leadership to bring agencies together in pursuit of a common solution. As a result, agencies like IRS, DOJ, DoD, DOI and others are all struggling with limited resources to solve the very same problems. Sol concluded by noting that IRS has been learning from the work of others and agrees with the recommendation for a distributed set of federated registries contained in the business case.

Sol’s presentation is available at http://xml.gov/presentations/irs1/xmlinitiatives_files/frame.htm

Felix Martinez of BearingPoint and David McDonald of Microsoft demonstrated the Enterprise Planning, Investment Control & Collaboration (EPICCS) application BearingPoint has developed running on SharePoint and InfoPath. David said 18 agencies used InfoPath to submit Exhibit 300s to OMB last year and that a collaboration Web site is now available to provide for simple workflow, security, versioning, and check-in/check-out of documents. Role-based security is built into SharePoint.

Each business case has its own URL but is broken up into sections that can be processed separately by different people. Guidance appears in the right pane in InfoPath, including a library of examples that have been rated highly by OMB. David said they wrote an XSD that is separate from the form itself for context-sensitive help tailored to the customer. Felix noted that drop down validation/lookup tables and spreadsheet capabilities are provided, taking advantage of the richness of InfoPath. The document library check-in/check-out, versioning and history, and event handler features of SharePoint. The library of XML files can be reused in the next budget cycle, and David noted that InfoPath can automatically update the Exhibit 300 form when OMB changes the XSD. Ken Sall asked about large changes and David said they could be handled by writing custom code. Finally, Felix said many customers already have many of the Microsoft components required and can upgrade to take advantage of EPICCS at very low cost.

Felix and David demonstrated some of the features of EPICCS. They did not have a presentation per se but information on the application is available at http://www.bearingpoint.com/epiccs/thank%5Fyou.html

Martin Henning briefed the group on FDIC’s project to apply Extensible Business Reporting Language (XBRL) to their Call Report. The report is very standardized not only in terms of data but also presentation. It contains about 2,000 data fields comprising information from the balance sheets and income statements. Four hundred page of instructions are provided and 1500 edit formulas are required to compile the reports. It is a very laborious process and is largely handled by phone, which is why it is called the “Call Report”. Seven vendors currently provide software to support the process, via spreadsheets and PDFs with instructions.

The Central Data Repository (CDR) will use XBRL taxonomies for content, instructions, and validation criteria. The reports will be received and published as XBRL instance documents. Many companies are repurposing the data but the lag time for some of those purposes may be as long as two months, whereas XBRL will facilitate reuse of the data in near-real time. Both the presentation as well as the data itself are being addressed in XBRL. Business rules are also being expressed in XBRL.

John Wizneiski said the project is testing with one dataset but is being built modularly so that its components can be reused with SEC, IRS and other financial reports. XBRL is a global, rather than just a U.S. standard, and it must accommodate two different accounting standards – GAAP in the U.S. and IAS internationally. Joe Chiusano asked if any thought has been given to using OWL in the future, in lieu of XLink and Linkbase. Judith Newton asked how FDIC has achieved agreement on the element names and has established linkages to the names being used. John said the taxonomy is a series of linkbases. The instructions currently comprise 400 pages in a Word document, but FDIC plans to put them in a database to facilitate updating and searching. They are trying to bridge from forms to concepts.

In the future banks will make their submissions via the public Internet rather than by virtual private networks, as they do now. Unisys is identifying standards that are relevant to the project, and FDIC is aiming to facilitate reuse, e.g., by the insurance industry. However, XBRL needs more validation capabilities (e.g., calculations) and agreeing on standards is a big challenge, particularly with respect to definitions and mappings to other environments. In response to a question, John indicated deployment of the system is planned for October 1, 2004. He said very granular and mature data is available in the financial industry, but it needs more work in terms of mapping and linking to taxonomies.

Martin’s presentation is available at http://xml.gov/presentations/fdic1/XBRLNewCall_files/frame.htm

Those in physical attendance included:

Owen Ambur, Co-Chair

Roy Morgan, NIST

Judith Newton, NIST

Susie Adams, Microsoft

Richard Campbell, FDIC

Joe Chiusano, Booz Allen Hamilton

David Fritz, Hummingbird

Martin Henning, FDIC

David Hoelzer, Xlipstream

Steve Jacek, PureEdge

Joab Jackson, Government Computer News

Dave Kamienski, PureEdge

Kirk Markland, Plexus Scientific Corporation

David McDonald, Micosoft

Kathleen Morgan, SCS/IRS-Prime

Vicky Niblett, SAIC/NASA

Greg O’Connell, PureEdge

Sue Reber, Plexus Scientific Corporation

Lisa Ruff, Microsoft

Russ Ruggiero, HumanML

Sol Safran, IRS

Kenneth Sall, Silosmashers

John Sarazan, Level 8 Systems

John Wizneiski, FDIC

Participating by teleconference were:

Robert Benedict, NASA

Rex Brooks, OASIS, HumanML & WSRP TCs

Please convey any additions or corrections to Owen_Ambur@ios.doi.gov