XML Community of Practice
Meeting Notes
August 16, 2006
This meeting was hosted by IBM at its office on K Street in Washington, DC.
Attendees and teleconference participants introduced themselves. Owen Ambur thanked IBM for hosting and made a few announcements. He noted that the XBRL CoP was scheduled to meet the next day and that the XML Schema Interoperability work group (XSI WG) teleconference was scheduled to occur at the same time. He mentioned that the September 19 Collaborative Expedition Workshop will focus on XBRL and the October 24 workshop will occur the day before the 10th anniversary of Raines’ Rules and may focus on lessons learned since they were issued. Finally, he noted that Terry Brady would be making a presentation on XForms at the DC Area XML User Group meeting in the evening (i.e., on August 16).
Following the announcements, with the assistance of Greg O’Connell, Owen displayed and briefly discussed the ET.gov site, pointing out the new submissions, XML-related submissions, and Stage 2 and 3 CoPs.
Mike Rowling, Darren Peterson, and Steve Hord briefed the group on Xforms, XFDL, and XBRL and demonstrated the use of those technical specifications with Census form F-72, pertaining to the reporting of State tax receipts.
Mike noted that XFDL deals with the presentation of forms and is pure XML. He noted that EDI is too expensive for small organizations to use, but by using forms on one end of a process, EDI transactions can be delivered on the other end. With reference to record-keeping, he pointed out that when faxes were the primary means of transmitting data, the paper faxes themselves were kept as records after the data was entered into databases. Now when the data is transmitted electronically end-to-end, the presentation of the record is lost but XFDL can provide for the preservation of records with integrity of presentation, including the use of digital signatures. Owen asked if any users of IBM/PureEdge Workplace Forms are using the eGov eAuthentication service, and Greg indicated that Grants.gov is doing so.
Besides the client plugin application, Mike noted that Workplace Forms can also be delivered automatically via HTML and Javascript generated on the server, without the need for any additional integration or design work. He also said the forms designer has the capability to scan existing paper forms to provide for the fidelity of presentation of the forms in print and on computer screens. If a data model is available, he indicated the form can automatically be generated from it. In that regard, Owen pointed out that the Federal Enterprise Architecture (FEA) Data Reference Model (DRM) is supposed to become the governmentwide data model, and that it will be good when forms can automatically be generated from the DRM.
Following the mid-morning break, Steve Hord noted that Charlie Hoffman received a grant to start specifying an XML schema for financial statements but quickly encountered some weaknesses in XML Schema. The XBRL specification arose from his initiative to address those weaknesses for the purpose of financial reporting, for a use case involving semi-structured data. Key features of XBRL include:
● A master dictionary of terms
● A highly normalized data model
● Strict extensibility rules
● A context validation rules engine
Business rules are modeled right in the schema. XBRL is a data modeling language. It models semantics, and as many different views of the data can be generated as necessary. In addition, whereas XML Schema requires only one root element, XBRL enables use of as many root elements as desired. XBRL can be embedded within XForms, and in response to a question, Steve acknowledged that it becomes pure XML at that point. XSLT can be used to transform the data for different views and purposes. UBmatrix is striving to make XBRL easy to use by anyone without requiring any knowledge of XML. Steve demonstrated how an XForms form can automatically be created from an XBRL data model.
John Triplett asked if XForms is a subset of XBRL and Steve explained that forms software is based on a page metaphor whereas XBRL is data model centric, providing a holistic view of the model. Mike noted that besides the data model, the presentation layer can be reused as well. For example, a set of address fields can be reused to appear the same way on multiple forms. Mark Montoya asked where the business rules are represented. Steve responded that they can be included in the data model but that you can “have it your way” by starting wherever you can. Darren noted that XForms forms can be routed outside the process and Mike added that XForms is an open standard format, as contrasted to proprietary formats like Exel.
Mike, Darren, and Steve’s presentation is available at http://xml.gov/presentations.asp#20060816
Keith Gardenhire briefed the group on DB2 version 9, particularly the XML related features. He noted that XQuery does not yet provide a standard for updating the database so IBM has provided proprietary means for doing so. DB2 also provides for the compression of XML in the database, and an index can be created on any element in the database. Currently, DB2 supports XQuery 1.0 and only well-formed XML is supported. Row- or column-level security is supported, as determined by the database security administrator. John Triplett asked if that means security can only be controlled at the document and not the element level, and Keith replied, yes, the database row is the lowest level at which access can currently be controlled. John also asked about the scalability of native XML databases.
Keith did not make his presentation available for posting on the xml.gov site. However, Brian Williams made a presentation on DB2 at the DC Area XML User Group meeting on September 20.
Those who registered their presence at this meeting included:
Owen Ambur, Co-Chair, xmlCoP
Richard Campbell, FDIC
Tom Cowell, IBM
Keith Gardenhire, IBM
Karl Hebenstreit, GSA
Steve Hord, UBmatrix
Jackie Kleinbelder, IRS
Bill Levis, Urban Institute
Glen Mazza, DoD
Mark Montoya, FDIC
Greg O’Connell, IBM
Darren Peterson, UBmatrix
Mike Rowling, IBM
John Triplett, IRS
Daniel Varhese, DoS
Zena Washington, IBM
Henry Wulf, Census
Michael Yamener, Census
Those who identified themselves as participating via teleconference were:
Jared Andrews, LMI
Allyson Ugarte, XBRL US & Spain (dialing in from Spain)
Quyen Nguyen, NARA, tried to dial in but was unable to get through. However, he did receive the presentations E-mailed to him just prior to the meeting.
Please convey any additions or corrections to Owen_Ambur@ios.doi.gov