eXtensible Markup Language (XML):
Greek, Esperanto, Panacea or Snake Oil?

Owen Ambur, Division of Information Resources Management (1)
U.S. Fish and Wildlife Service, Department of the Interior
June 9, 2000

Where the Java language provides for portable code,
XML provides for portable data.(2)

XML is smart data. HTML tells you how the data should look,
but XML tells you what it means.(3)

Another acronym. Just what we need. They're all Greek to most of us anyway, and since they've solved all our problems, why not another one? ...

Kidding aside, when it comes to technology and new technical standards, a healthy dose of skepticism is not bad medicine. However, even after filtering out the usual degree of hype, the eXtensible Markup Language (XML) seems to be the real deal. One way or another, we're going to be hearing more about it. The issue is what we may want to think about doing with it, bearing in mind our goals and objectives in pursuit of our statutory missions.

In essence the mission of the U.S. Fish and Wildlife Service is to promote biodiversity. In nature and by nature, diversity promotes strength in ecosystems so that they can endure and survive adversity. Likewise, cultural diversity adds zest, vitality, and strength to our communities, to the Department of the Interior, and to our nation as a whole. However, biological diversity is grounded in the laws of nature and cultural diversity is embraced by the principles of basic human rights. In short, it is the existence of objective standards that both undergirds natural systems and releases the creativity of the human spirit.

XML is an information technology (IT) standard issued by the World Wide Web Consortium (W3C) in 1998. It is one of a family of "markup languages" fashioned to provide a standardized, non-proprietary means of structuring and exchanging information.(4) It was designed to provide a simplified "grammar" upon which organizations worldwide can base vocabularies that have application in their industry or domain. The promise of XML is to enable organizations to automate the exchange of information without the need for translation or conversion.(5) In other words, XML embodies the potential to overcome the Tower of Babel that so commonly characterizes information systems in large organizations and thus afflicts individuals who are called upon to use them.

Reflecting that visionary promise, what are some practical ways in which we in the Department of the Interior may wish to capitalize on the potential of XML? Given its youth and the indefinite scope of its ultimate breadth, it is impossible at this point to envision all of the possibilities that will become apparent as the standard matures in the marketplace. However, in the spirit of discovery, here are some suggestions that may warrant exploration, grouped into eight general categories and with some of the driving forces noted, as follows:

  1. Directory Services - Customer-focus, privacy, security, extranets, virtual private networks, digital signatures, Paperwork Reduction Act (GILS and information inventory), White and Blue Pages, quality management(6)
  2. Forms Automation - Government Paperwork Elimination Act (GPEA), President's E-Gov directive, data lookup, validation, and reuse(7)
  3. Data Standards and Architecture - Interoperability, elimination of proprietary stovepipe systems that become isolated islands of information, emerging consensus(8)
  4. Workflow Automation - Reinvention of government, process streamlining, information collection burden reduction(9)
  5. Document/Records Management - Electronic Freedom of Information Act Amendments (E-FOIA), subpoenas and discovery in litigation, efficient reuse of information in internal business processes (knowledge management), risk reduction(10)
  6. Access and Accessibility - Sec. 508 of the Rehabilitation Act, fairness, desire to "do the right thing," potential for lawsuits(11)
  7. Reports Automation - Reuse of data, financial reporting and auditing, Government Performance and Results Act (GPRA), the Internet(12)
  8. Information Translation Services - E-commerce, E-government, affirmative action for small and disadvantaged businesses, economic efficiencies for big businesses(13)
Directory Services - Much lip service is paid to the notion of customer focus, but the reality is that people are often treated as afterthoughts in the design of IT systems. Even when users are directly and intimately involved from the outset, the focus is on a narrowly defined set of requirements and a select group of the people who will be expected to interact with the system in some fashion. The so-called "user interface" is often the last piece of the puzzle, developed months if not years after the specialized "back end" features have been put in place.

The results are stovepipe systems that treat users as properties to be acquired, rather than as human beings to be served. In the many-to-many problem of systems design, it is the people who are expected to yield and dance to the favorite tunes of myriad systems. Such are the inevitable results of the failure to use a standards-based approach to systems development - focusing first and foremost on people as people, rather than as appendages to any particular system. By definition, such systems are proprietary to their "hosts" rather than being customer-focused.

While any promotional material should be taken with a grain of salt, here's what Novell says about its XML-based directory product:

DirXML seamlessly links and manages all places where user profile information is stored, including software applications, network operating systems, databases and network devices. DirXML is the first directory-based technology that not only brings all data together, but also keeps control of information with the individual departments that own the data. This enables organizations to distribute data ownership while centralizing information availability.(14)
Novell, Microsoft, IBM, Oracle, and the Sun-Netscape Alliance are all supporting Directory Services Markup Language (DSML), which is touted as "the XML Standard for Directories that Enables Friction-free eCommerce."(15)

That sounds good. However, regardless of the merits of any particular product or implementation, it is clear that standards like XML, X.500, X.509 and LDAP are indispensable parts of any solution that provides the most important kind of "interoperability" - the operability of many information systems by many different individual people!(16)

A single logical, distributed, standards-based person directory is a very good place to start all IT systems development projects - especially in an organization whose missions and stakeholder groups are as large and diverse as the Department of the Interior. While systems, objectives, goals, and even missions change, people remain people. Serving their needs should be "job one" of every taxpayer-funded system. Are we or are we not are a government "of, by and for the people"?

Trusting that we are, another appropriate focus for early and comprehensive attention is the definition of our missions, goals, and objectives in terms that are commonly understood by our external stakeholders. Agencies are required by the Government Performance and Results Act (GPRA) to document their goals and objectives, to link them to the inputs they receive, and to report the results. Meanwhile, under the Vice President's leadership and direction of the National Partnership for Reinventing Government (NPRG), agencies have been striving to improve their listings in the so-called Blue Pages of the telephone directories. Thus, the potential exists not only to identify our telephone numbers by commonly understood functional terms but also our Web page URLs, E-mail addresses, E-records series and, ultimately, individual E-records and even discrete elements within each of our E-records, if appropriate.

In this context, it should also be noted that the Paperwork Reduction Act (PRA) requires each agency to compile and maintain a "current and complete inventory" of its information resources.(17) In closely related provisions, the Electronic Freedom of Information Act Amendments (E-FOIA) require agencies to: a) provide an online index of frequently requested records, and b) make reasonable efforts to search for records by electronic means.(18) One of the principles of Total Quality Management (TQM) is that we should aim not merely to satisfy but to delight our customers. In pursuit of that objective, why shouldn't we provide an online index not only of our frequently requested records but all of our records - especially since most of them are already routinely generated and processed by electronic means?(19) (See also Access and Accessibility, below.)

GPRA documentation would be an especially high-value set of records series to be comprehensively indexed for access by our stakeholders. (See also Reports Automation, below.) XML provides a standardized, nonproprietary means by which to facilitate the compilation and use of such a stakeholder-focused, results-oriented directory service.

Forms Automation - Since the user interface is often the last feature addressed in systems development projects, forms are frequently viewed primarily as extensions of databases rather than representations of people. However, a true customer-focus leads to recognition of the fact that forms are data surrogates for the people who are called upon to use them.

Again, the many-to-many problem is key. Should individual people be forced to render themselves in myriad forms automation software programs and systems? Conversely, shouldn't it be possible to render any form in any standards-compliant client software that the individual may choose to use?(20) Once again XML is a key enabler. In particular, by clearly demarcating and defining individual elements within forms (or for that matter, any document), XML can be used to facilitate the exchange of data among myriad client and host applications.(21)

Ultimately, any individual should be able to use a single software client program to process any electronic form and seamlessly transmit the data to any database, subject only to valid rights to do so, anywhere in the world. Indeed, that is the aim of the W3C's XForms Working Group, which succinctly describes XForms as "the next generation of Web forms."(22) Key objectives for the XForms group include:

The Government Paperwork Elimination Act (GPEA) requires agencies to make their forms available by electronic means no later than October 2003.(23) Moreover, the second of the President's E-Gov directives mandates that the forms needed for the top 500 Government services be made available online by the end of calendar year 2000.(24) While those services have yet to be expressly identified, a good goal would be for all agencies to act as if all of their services are among the top 500. Accordingly, a key objective for agencies might be to make all of their forms available on line in intelligent XML (XForms) format. Ultimately, the intelligence built into each form should include not only the constraints imposed upon each data element (e.g., character type and element length) but also the values that may be inserted in each field (e.g., by referencing a validation/lookup table). (See XML Schema, below.)

Data Standards and Architecture - The establishment of data standards and particularly the delineation of IT systems architectures are sometimes viewed as ivory tower exercises that are either benign or pernicious to those who are in positions to make actionable decisions on specific projects and systems.(25) In particular, standards for electronic data interchange (EDI) have been viewed as too inflexible, costly, and cumbersome for any but the largest organizations to implement.(26) That is an understandable human reaction to activities for which no immediate benefit is apparent and over which the individuals in question have no direct control. However, XML holds the potential to change the dynamic and the incentive, by making the benefits of the establishment of standards directly and immediately available on the Internet.(27)

Indeed, XML itself is a "data architecture" upon which IT systems can and are being built, at breakneck speed. Those who have little use for architectural planning essentially have two choices: 1) sit back and do nothing while XML washes over them, or 2) get with it and help to define each and every data element and schema that should be uniquely described in XML. Likewise, those who are presently engaged in IT architectural planning have the two choices: 1) continue laboring in obscurity in their ivory towers, or 2) start rendering the fruits of their labors in readily accessible XML elements and schemas on the Internet. In both instances for both groups, the results will speak for themselves. The only issue is the relevance of their roles in the revolution that is XML, and the evolution of its many facets. XML offers the prospect of a win/win proposition for all concerned, by capitalizing on existing Internet standards to dramatically shrink the gap between theory and practice.

If we are truly serious about taking a customer focus, the initial thrust of data and systems architectural efforts should be to define the elements required to describe the people (including the roles they play) in our work processes, and to make those elements readily available in a single, logical, distributed, standards-compliant directory. All subsequent elements should be delineated and systems designed around the constellation of elements comprising the person and roles directory.(28)

And what are the architectural components of XML documents themselves? Each document is comprised of a series of logical elements delimited by markers called "tags" and each element may have subordinate (nested) elements.Attributes (sometimes called "properties") may be appended to each tag to identify how to process each element. They may describe the characteristics of the element, including reference to external standards. Together with their associated attributes, XML tags are equivalent to a data dictionary. XML may be transmitted as an unformatted data stream or it may be combined with a stylesheet as part of an HTML file for presentation in the user's device.

Normally, the "root" is the first element, within which all of the other elements are nested, and the name of the root element matches a Document Type Definition (DTD). Each DTD is essentially an architecture for the class of documents that it describes.(29) More explicitly:

A DTD provides a formal description of the full structure of an XML document with relevant sets of values. This description includes the ordering (looping), the sequence, and the relationships of all tags (elements) contained in the document. A DTD is not mandatory. If it is not present, the XML system creates a default definition for all unidentified tags. However, XML documents that do not contain or reference a DTD must adhere to stricter XML specifications and are limited with respect to processing capabilities.(30)
A DTD need only be transmitted only once. Subsequent exchanges of data can reference a copy of the original DTD stored locally. Alternatively, DTDs can reside at remote locations (such as a centralized repository of standard, commonly used DTDs) and be referenced in the header of XML documents. DTDs have obvious potential to facilitate delivery, presentation, and use of the latest approved versions of Standard and Optional government forms. However, as with any technology, the limitations of DTDs are already becoming apparent and a recommendation has been proposed to begin to address them:
XML Schema - The XML schema specification is being created to develop more powerful XML-based DTDs. Schemas provide rules for validating business transactions and can be used to define those transactions with particular stakeholder groups. There are two parts to the specification:
Part 1. Structures - Methods for describing the structure and constraining the content of an XML document.

Part 2. Data Types - A framework for associating data types with XML element types and attributes, enabling software to manage dates, numbers, and other special types of data more efficiently.

To the greatest degree possible, government agencies should use schemata that are common in the private sector, preferably those that are accepted as international, world wide standards.(31) However, to the extent that "we the people of the United States of America" do in fact have unique data requirements, including requirements that are not being effectively met by industry standards, agency decision-makers have an obligation to exert leadership in the establishment of such standards.(32) Specifically, they are charged with fostering the development and use of voluntary consensus standards, rather than attempting to dictate usage of arbitrary or proprietary standards.(33)

Workflow Automation - The Workflow Management Coalition (WfMC) has issued an XML-based specification to provide for interoperability among workflow management systems. The intention is to describe a language that can be used to achieve two basic types of interoperability - simple chained workflows and nested workflows. The specification supports both types of interchange synchronously as well as asynchronously. It is independent of any particular implementation mechanism, such as programming language, data transport mechanism, platform, or hardware. Since one of the goals is to define a "light," easy-to-implement protocol, the standard focuses on the common aspects of workflow implementations and describes the data interchange in XML.(34)

To the degree that human beings are already overburdened with work, automation of the flow may add little, if any value. Indeed, the value may be negative. In the cyberage, information overload is a far greater problem than the lack of "flow" of information.(35) However, to the extent that high volume work products follow a routine and predictable pattern, automation can relieve humans of the burden of facilitating and tracking such flows. That is particularly true if such work products are already in well-formed XML documents that can be readily interpreted and processed by machines without human intervention.

The combination of workflow automation with XForms, DTDs, and XML Schema to ensure well-formatted transaction sets - together with XML-based person and roles directories to facilitate data flows - could be potent indeed. Most importantly, it can relieve human beings of the tedium of routine processing requirements and thereby free their creative energies to focus on more complex, higher-value, and more personally rewarding tasks.(36)

The key issue for workflow automation is whether any particular application is user-centric or process-centric. In other words, does it serve the needs and wishes of the people involved in the process, or does it merely serve to enslave them to the process?(37)

Document/Records Management - XML supports both a user-centric as well as a document-centric model. That is, discrete data elements are derived in the context (document) within which they occur.(38) At the same time, well-structured tagging facilitates the automated parsing of discrete data elements for exchange with databases and other applications. From a records management perspective, that enables us to have our cake and eat it too. That is, we can efficiently and dynamically manipulate, process, relate, and re-relate data elements to each other while at the same time preserving as inviolate the original source documentation (i.e., the record) from which each element derives. That means not only that audits can be largely automated but also that records can be appropriately preserved for historical reference without impeding dynamic, current work processes.

Properly employed, XML metatags can also facilitate the search and retrieval of records during the active and passive stages of their life cycles, as well as automated destruction at the appropriately scheduled time.

The most important reason for organizations and individuals to engage in sound records management practices is to enhance the efficiency and effectiveness of their own business processes. Traditionally, the focus of records managers has been twofold: 1) usage of records for historical purposes, and 2) determining when boxes of paper records can be discarded or shipped to the archives in order to save space, reduce clutter, and make the remaining records more readily accessible within the organization. Unfortunately, all too often, litigation or the threat of litigation has been the primary motivational factor - a risk that has been greatly magnified by the persistence of the plethora of uncontrolled, unmanaged E-records that commonly accompany the use of IT.(39)

Documents embody knowledge as best it can be represented by those who make it explicit, i.e., record it. Metadata makes explicit the relationships that are implicit in the knowledge contained within any record.(40) XML can help make the metadata and, thus, both the relationships as well as the records themselves readily discoverable and available for reuse on the Internet. In turn, ease of access to the knowledge documented in E-records will stimulate the discovery and advancement of new relationships among people and ideas in pursuit of the ever-expanding base of explicit knowledge.

Access and Accessibility - Seldom has public access, including accessibility by disabled persons, been a significant factor in how organizations treat their records. Indeed, accessibility of records within the organization itself is commonly paid far too little heed by decision-makers and systems designers.(41) It is almost as if the common outputs of all IT systems (i.e., E-records) are considered to have zero value once they have departed the screen of a computer monitor and/or been sent to a printer ... even though they remain in more or, more probably, less well-ordered fashion on computer hard-drives and backup tapes.

Accessibility occurs on two levels - machine-centric and user-centric. A record is considered to be accessible if it can be equally understood by its targeted audience regardless of the device used to access it. While the technicians have placed the emphasis on multi-modality (device independence), the semantics (meaning) of the elements are equally important to the users (people). Without knowledge of the meaning of the elements and attributes, chances are diminished that user software agents can do anything intelligent with documents. Semantics knowledge can be provided through human readable documentation, but machine readable assertions of some semantics are important to enable the rendition of documents in various media. As one commentary notes, semantics are "... paramount for pervasive access" so that "... you don't need a programmer ... just ... a program."(42)

Documents that are rendered in "well formed" XML will be "readable" not only by commonly used devices but also by assistive technologies for those whose physical capabilities are diminished to various degrees. More importantly, well-structured XML documents will be understandable by people, regardless of the devices they use to render the knowledge contained therein. Disability is a relative term. We are all "disabled" in varying degrees. Technology is the means we use to overcome our human frailties. XML is a standard technology that embodies a powerful, indeed a wonderful potential to enable all of us to participate more fully and effectively as human beings in the progressive pursuit of knowledge.(43) Naturally, those who are fair minded and public spirited will want to make our records readily available to all, regardless of physical disabilities. However, the natural inclination of agency leaders to provide the necessary direction is bolstered by Section 508 of the Rehabilitation Act, as amended by the Workforce Investment Act of 1998, under which agencies may be sued if they fail to make their IT applications accessible.

A vital aspect of accessibility is the ability to find and retrieve E-records when and where they are needed. Various sets of metadata have been promoted to facilitate the discovery of information.(44) However, lack of consensus on protocols, procedures, and widely accepted tools for the creation and use of metadata has hampered implementation. A Catch 22 has resulted: The lack of reasonably comprehensive, up-to-date, readily available sets of E-records metadata discourages potential query-users, while the creators of E-records lack the incentive to support repositories of metadata in absence of a clear demand for it.(45) XML holds the potential to break that vicious cycle by establishing a widely accepted, flexible, and easy to use means of generating and making metadata available over the Internet.(46)

Reports Automation - The Extensible Business Reporting Language (XBRL) is an open specification that uses XML-based data tags to describe financial statements for both public and private companies.(47) XBRL is not transaction oriented. That is, it is not designed to facilitate the creation and transmission of well-formed XML-based transaction sets or documents. Instead, it is designed to glean and redisplay in various financial reports the pertinent elements that have been appropriately tagged in various other types of XML documents. In particular, it aims to give auditors and analysts, including arm-chair investors, the financial data they need in a format that is both readily comprehensible as well as easy to manipulate and use.(48)

While the initial focus of XBRL is on financial reports that companies are required to file for regulatory purposes, an interest group is also being formed to address financial reporting requirements that are internal to government agencies themselves. Moreover, many of the same features can be used to report non-financial quantitative and even qualitative performance measures. The implications for the reporting of GPRA performance measures are obvious. In the near future, near-real time reporting on the Internet will not only be expected, it will be routine. Agency stakeholders have every right to expect nothing less.

Indeed, to the degree that original source E-records are made available on the Internet in well formed XML that complies with applicable schema standards, many current reporting requirements will become unnecessary. Anyone who chooses to do so will be able to select the pertinent elements and generate whatever reports they desire at any time. Of course, too, they will be able to use software agents to monitor the Net and deliver reports to them whenever any change occurs to a pertinent parameter. Opportunities for aggregators, value-added resellers, and other intermediaries will abound. However, unlike intermediaries of the past, they will be adding real value to the information itself, rather than merely bridging gaps in inefficient supply chains based upon paper-based and word-of-mouth processes. In terms of customer focus, more suppliers will be able to address the needs and wishes of customers directly, without having to work through needless layers of administrative overhead.

Translation Services - One of the beauties of eXtensible Markup Language is that it is extensible - meaning that those who have the need or desire to do so can easily extend it to encompass features that are not addressed in other implementations. At the same time, any implementation that complies with the XML standard will create E-records usable in large measure by any other XML-compliant application. Some of the initiatives that will provide the necessary translation services include the following(49):

Resource Description Framework (RDF) has been described as "the equivalent of a master library catalog index card for data on the Web."
RDF is important to facilitate data exchanges and searches for relevant information.(50) In particular, it will help to document the meaning (semantics) of data and metadata elements, thereby facilitating the development of thesauri and other finding aids. One potential application is the Blue Pages directory service, whereby government functions could both be related to and distinguished from each other. Another would be the classification of E-records not only for retention but especially for search and retrieval purposes, including responses to FOIA requests, subpoenas, and discovery in litigation.(51)
XSL Transformations (XSLT) provides the syntax and semantics for changing one XML document into another, using source and destination document trees.(52)
One potential use of XSLT, for example, might be to take the pertinent data from a 300B Exhibit (Capital Asset Plan and Justification) for a major IT project and redisplay selected elements in the context of a summary for review by an agency's information technology investment review board. Another possibility would be to use selected data in Exhibit 53, which summarizes the budget figures for all of an agency's IT projects, to automatically render Exhibit 52, which pertains to the subset of IT projects that include financial management components.(53)
Extensible HyperText Markup Languge (XHTML) is a reformulation of HTML 4.0 in XML.
XHTML integrates XML capabilities into HTML to create standards that support new Web platforms, e.g., cell phones, televisions, automobiles, small wireless devices, kiosks, and nontraditional PC desktops.

Finally, it would be an oversight not to mention an important international initiative known as Electronic Business XML (ebXML), the mission of which is:

To provide an open XML-based infrastructure enabling the global use of electronic business information in an interoperable, secure and consistent manner by all parties.
ebXML is a joint effort of the United Nations body for Trade Facilitation and Electronic Business (UN/CEFACT) and the Organization for the Advancement of Structured Information Standards (OASIS). A primary objective is to lower the barrier of entry to electronic business in order to facilitate trade, particularly with respect to small- and medium-sized enterprises (SMEs) and developing nations.(54)

Summing Up - So there you have it - XML, the solution to all of our problems... Oh, really?... No, not by a long shot.(55) Nor is this a comprehensive listing of all of the related initiatives, much less the myriad practical issues involved. For example, who's going to do all that "tagging"? Even if some of us do take a crack at it with some of our files, what's to say we'll do it right? What's the payoff for us? What assurances do we have that proprietary extensions pushed by Microsoft and others won't defeat the purpose? And, by the way, where are all of those whiz bang "user interface" tools to assist us with the tagging?(56) Even if we choose one of the rudimentary early candidates, won't people be dissatisfied with them? Won't a better bunch of tools come along soon? Why not just wait?(57)

These are good questions and there are many others that may be cause for pause. XML may provide for "smart" and "portable" data but it is no panacea. Its limitations will only become more apparent as we become more familiar with it as a standard. Yet that is the nature of the unrelenting march of knowledge. Despite its imperfections and the expectations it imposes upon us as human beings, the potential of the knowledge already embedded in the XML standard is truly vast.(58) If we choose not to participate wholeheartedly in the quest to capitalize on it, we can rest assured that others will.(59)

The future will be what each and every one of us - individually and collectively - chooses to make of it. XML is nothing more ... nor nothing less ... than an open-systems, standards-based foundation upon which we can work more effectively together to build a brighter tomorrow ... even as we celebrate our differences.

End Notes

1. Owen Ambur is a Staff Assistant (Systems Analyst) in the Division of Information Resources Management of the U.S. Fish and Wildlife Service. Together with Martin Smith of the U.S. International Trade Commission (USITC), he formed The Adhocracy on XML for Uncle Sam (taXus), which led to the chartering of the XML Working Group by the CIO Council's Enterprise Interoperability and Emerging IT Committee. Owen has been asked to co-chair the Working Group with Marion Royal of GSA's E-Commerce Office.

2. The quotation is from a comment by David Frietsch, Los Alamos National Lab, in an E-mail exchange with Carol Blackston of the Department of Energy, March 28, 2000.

Building on Mr. Frietsh's observation, it is noteworthy that the use of portable code increases the risk of receipt of malicious code. Such software programs may alter or destroy E-records without the knowledge of the user or her organization. Thus, the use of mobile code is potentially pernicious, if not antithetical to the effective management of information (E-records). Implicit in the use of mobile code is the notion that its "hosts" are better qualified to determine what programs (software) should run on the customer's computers than they or their own organizations are. Conversely, implicit in the concept of portable data is the notion that any client software that any customer may choose should be able to read and interact with any E-record without being forced to run software programming logic that is proprietary to its "host".

Portable code supports the "push" paradigm, while portable data supports the "pull" paradigm. The push paradigm is designed to serve the interests of suppliers ("hosts") whereas the pull paradigm is customer (client) focused.

3. This quotation is from "XML in An Instant: A Non-geeky Introduction" by Charles Goldfarb at http://www.oasis-open.org/html/goldfarb.htm

4. Other examples include Hypertext Markup Language (HTML), which is the language of the World Wide Web, and Standard Generalized Markup Language (SGML), which is a superset of XML.  Additional information on markup languages, particularly XML and SGML, is available in the XML Cover Pages, at http://www.oasis-open.org/cover/

Larry Mongin provides a primer on XML at http://www.bloomington.in.us/~mongin/xml-tutorial/ XML FAQ are at http://www.ucc.ie/xml/. A series of white papers is available at http://www.xml.org/xmlorg_resources/whitepapers.shtml. Sun Microsystems provides "A Quick Introduction to XML" at http://java.sun.com/xml/docs/tutorial/overview/1_xml.html

5. Paraphrased from white paper provided by Barclay Blair of PureEdge in an E-mail message, May 11, 2000.

6. The importance of privacy as a public issue can be measured in the number of bills that have been introduced in Congress as well as the number of articles that have appeared in the press. Multinational companies have feared that the European privacy directive could shut down international commerce. Security is viewed as "the next big thing" following on the Y2K scramble. Organizations that have developed Internet and intranet sites continue to struggle with the needless overhead associated with conducting secure and authenticated communications with external stakeholders. Standards-based extranets, virtual private networks, and digital signatures all have vital roles to play.

The Paperwork Reduction Act requires each agency to index its major information systems, holdings, and dissemination products in the Government Information Locator Service (GILS) and to maintain a "current and complete inventory" of their information resources. [44 USC 3511 and 3506(b)(4)]

Under the auspices of the CIO Council, the Directory Forum is pursuing implementation of a Governmentwide White Pages (person directory) service. Under the auspices of the National Partnership for Reinventing Government (NPRG) and the leadership of the Vice President, agencies are working to improve their listings in the Blue Pages (government functions and services). Finally, while the results may call into question the commitment of agencies to quality management, few will question the need for it.

7. The Government Paperwork Elimination Act (GPEA) requires agencies to make their public-use forms available for online completion, including the use of digital signatures, by October 2003. The President's E-Gov directive requires that the forms required for the top 500 Government services be online by the end of calendar year 2000. To avoid needless data entry errors, lookup and validation services. To avoid needless rekeying of data, standard-based means must be provided to enable its reuse wherever and whenever there is a valid need to do so.

8. As people and organizations mature in their use of information technology, recognition is growing that interoperability is key. The time is growing short when customers, suppliers, individual citizens, employees, and organizational leaders will tolerate the inefficiencies and burdens imposed by proprietary systems. Congress need not pass any more laws telling agencies to refrain from acquiring and developing stovepipe systems, which inevitably become isolated islands of information. Consensus is emerging that such systems should be eliminated as rapidly as possible.

9. While bureaucratic organizations are slow to embrace change, reinvention of government is moving beyond the slogan stage. Regardless of who sits in the White House, the reality is setting in that "the only constant is change." Public servants owe it to the taxpayers to streamline any and all processes involved with the ever-expanding roles of government. The Paperwork Reduction Act requires agencies to minimize the burden of information collection requirements upon the public.

10. The Electronic Freedom of Information Act Amendments (E-FOIA) require agencies to store, search, and make their records available by electronic means. [5 USC 552(a)(2) and (3)(C)] Subpoenas drive records searches after-the-fact and discovery is the most costly aspect of litigation. Such costs clearly highlight the need for more effective management of E-records during the active stages of their life cycles, so that they are more readily accessible during the passive stages. The most important reason to manage E-records effectively is to provide for the efficient reuse of information in internal business processes (i.e., knowledge management). However, risk reduction is drawing increasing attention in our litigious society. To avoid both the perception as well as the reality of fraud, very bright lines need to be drawn between the appropriate (routine) and inappropriate destruction of records.

11. Sec. 508 of the Rehabilitation Act, as amended by the Workforce Investment Act, requires agencies to make their IT applications, including their Web pages, accessible to persons with disabilities. Beyond the mandate expressed in the words of the statute, the desire for fairness and the natural inclination of public employees and decision-makers to "do the right thing" are strong motivators, and in this case, they are backed by the potential for lawsuits. Agencies that fail to make their E-records and IT applications accessible are subject to being sued by any citizen as well as their own employees. A Web site devoted to this topic has been established at http://www.section508.gov/

12. All reports are essentially restatements and supplementation of data that has been documented in other E-records. Thus, the need to continually reuse of data is a driving force for better means by which to do so. Such requirements are particularly acute for financial reporting and auditing. Moreover, the Government Performance and Results Act (GPRA) expressly requires agencies to document their goals, plans, and objectives, to report their results, and to link them to the inputs they receive. The need for standards-based linkages between planning, budgeting, logistical, and reporting systems is apparent. The standards that comprise the Internet yield a platform on which to build automated, anytime/anywhere reporting systems.

13. E-commerce and E-government are buzzwords today. Tomorrow, they will become the routine. The "E" will not only be assumed, it will be expected. In the meantime, affirmative action for small and disadvantaged businesses is a driving force not only for the U.S. Government but for the United Nations as well as for governments worldwide. Likewise, economic efficiency is a driving force for big businesses in the highly competitive, worldwide marketplace.

14. Novell Announces Open Beta of DirXML Technology: Links and Manages Business-Critical Data Across One Net. March 27, 2000 http://www.novell.com/press/archive/2000/03/pr00047.html

15. The DSML home page is at http://www.dsml.org/. A critique of DSML and the efforts of Microsoft and others to impose proprietary extensions upon us, by Edward R. Mortimer, is available at http://pages.cthome.net/iact/IQN/6-1999oct/iqn6-Fight_X500_Standard.html

16. Unfortunately, the continuing proliferation of proprietary system-centric user IDs and passwords is evidence that many organizations still do not have the interests of their customers truly at heart. It is also a sign of the "tragedy of the commons," whereby relatively larger costs are imposed upon a group of stakeholders because those costs are distributed, ignored, and/or unknown, whereas the cost of a more efficient allocation of resources is focused within the purview of the decision-maker.

17. The term "complete inventory" is a tautology (or pleonasm, if you prefer). A listing is not an inventory unless it is as complete as possible at the time it is compiled. When human nature is involved, imperfections (or "uncontrolled variance," if you prefer) are naturally interjected into any process (by design or by default). However, with automated systems, the natural tolerances for error are vastly reduced. Thus, the Congressional intent in including the word "complete" in the law appears to be aimed at limiting the probability that either human error or intent will be permitted to thwart the creation and maintenance of a "current inventory" - meaning a listing that is not only complete but also up-to-date. Moreover, in the cyberage, "up-to-date" might be taken to mean, up-to-nanosecond.

18. The pertinent provisions of the Paperwork Reduction Act and E-FOIA are 44 USC 3506(b)(4) and 5 USC 552(a)(2) and (3)(C), respectively.

19. E-FOIA also requires, within reason, that any record created after October 31, 1996, must be made available in electronic form upon request. Since most records are now created and processed by electronic means, it will become increasingly difficult to justify failure to make them available by such means.

20. Ideally, users should need nothing more than an XML-compliant Web browser - although the terms "browser," "Explorer" and "Navigator" fail to capture the essence of the program logic they embody as their functionality increases. Nor does the term "thin client" truly apply as the girth of these software programs expands to securely, efficiently, and effectively meet the requirements for processing and management of business-quality documents and data in a business-like manner.

Thin-client computing is a misnomer and the thought that all intelligence resides in the "host" is a relic of the mainframe era. The real issue is the same as it has always been - how best to distribute, employ, and leverage the intelligence distributed throughout the organization. In terms of software, that does not mean the client should be "thin"; it means that it should be appropriately sized to embody all of the programming logic that each individual needs to perform routine, recurring forms automation and document/records management tasks.

While it is entirely appropriate, indeed required, for agency decision-makers to use commercial off-the-shelf (COTS) applications, it is neither essential nor logical for Uncle Sam or for any individual decision-maker to default to Microsoft the determination of agency requirements. Indeed, we are charged by OMB Circular A-119 with actively participating in the development as well as the implementation and use of voluntary consensus standards, including those for client software applications. E-records management is a function that even Bill Gates is likely to acknowledge warrants greater attention.

21. The ability to discretely tag every element of every document does not necessarily mean that it should be done. The costs and benefits should be taken into account. Generally speaking, documents for which such treatment is warranted are already rendered as forms. However, where relatively unstructured text is routinely used on a recurring basis, opportunities should be explored to identify the discrete elements embodied in the text, so as to determine whether the narrative adds real value or whether it should be boiled down to a defined set of highly structured fields (i.e., a form).

The distinction between tagging for format (e.g., paragraph and data element delimiters) versus content (semantics) should also be noted. XML authoring tools should be able to automatically tag for formatting. Likewise, expert systems and artificial intelligence applications may be able to assist by automating semantic tagging. However, for some purposes, human effort may be the best way to add value through metatagging and the incremental costs and benefits should be assessed. Such analyses should include the costs of not indexing E-records in terms that are important for such purposes as: a) facilitating public access, b) responding to FOIA requests, c) complying with subpoenas, and d) ensuring that E-records are securely controlled and preserved for appropriate periods of time, as well as destroyed when appropriate.

A key issue is whether agencies should be permitted to fail to capture and manage, as explicit knowledge, contextual and other information generated and clearly known during the routine course of creating and processing E-records. Common wisdom holds that it is too expensive to capture such metadata. However, common wisdom is commonly wrong. Clearly, some amount of metadata is and will always be associated with every E-record; otherwise it cannot be accessed in the active memory of a computer, much less stored on a hard drive or other media. Thus, the real issue is how much metadata should be associated with each logical class of E-records. In that regard, common wisdom should be reassessed when XML authoring tools become widely available and when schemata for the classification of E-records series have been specified, taking into account public access and accessibility requirements.

22. The XForms Working Group's home page is at http://www.w3.org/MarkUp/Forms/ Among the sponsoring members of the XForms standard are Cardiff, JetForm, and PureEdge.

Cardiff approaches the problem as a value-added reseller of Adobe's Portable Document Format (PDF), for which they have developed a back-end processor to convert data from PDF Fillable forms into XML format. Through its proposed eXtensible Forms Archictecture (XFA) JetForm aims to build upon its strength in the forms automation marketplace, by maintaining the separation of the data from its presentation so as to facilitate the accessibility of its forms by any device. Conversely, through its eXtensible Forms Description Language (XFDL), PureEdge aims to fully address more stringent requirements that may apply to the acceptance of some forms-based data in a court of law, including perhaps the requirement that the data be inextricably bound to its presentation so that there is no doubt as to what the parties have seen and agreed.

However, despite their differing approaches, these companies and others have come together in the realization that a commonly accepted standard like XForms is necessary to reduce the risk and speed the adoption of their products in the marketplace. Now what remains is for decision-makers in the marketplace to realize their own interest in embracing the use of E-forms products that comply with the basic XForms standard and aim to comply with more fully specified standards, e.g., XFDL, in the near future.

23. A summary and the actual text of GPEA are available at: http://mysite.verizon.net/ambur/GPEA.htm

24. The President's E-Gov directive is available at: http://www.pub.whitehouse.gov/uri-res/I2R?urn:pdi://oma.eop.gov.us/1999/12/20/5.text.1

25. IT architectural efforts also typically spend too much time "running in place" while endeavoring to catalog the state of the world "as is," rather than simply expressing the state "to be" desired and leaving it to the creative genius of the minions to determine how best to transport themselves there. It is necessary to define the "as is" state only if the assumption is made that proprietary stovepipe systems will continue to be used, and the task is to figure out (e.g., by popular vote) which proprietary systems to use.

Conversely, if an open-systems, standards-based approach is used, all that is required is to invoke the pertinent standards and require that alternative approaches be expressly justified. (Alternatives may be either proprietary or based upon other standards.) XML is a standard with which noncompliance should be expressly justified. (So too is the DoD 5015.2 standard for E-records management.)

26. The XML/EDI group's home page is at http://www.xmledi-group.org/. The Data Interchange Standards Association (DISA) is involved with the XML/EDI group and makes the following assertions:

A foundation for e-commerce professionals and their organizations, DISA provides training and cross industry knowledge that can be easily accessed and applied to e-business requirements worldwide. Our presence not only offers educational venues to industry, but also encourages uniform implementations among trading partners.
DISA's home page is at http://www.disa.org/ (Note: disa.org should not be confused with disa.mil, which is the Defense Information Systems Agency, whose home page is at http://www.disa.mil/disahomejs.html. A search of disa.mil's site for the term "XML" does not currently pull up any records. However, that is a circumstance that does not seem likely to exist for long.)

27. The Internet is a network of networks. It is enabled by a group of standards that provide for interoperability of a defined set of services based upon an explicit set of technical specifications. While the inadequacies of those standards are apparent, particularly in terms of security, for example, the Internet could not exist without strict adherence to its technical specifications. Since those specifications are both explicit as well as "open" for use by anyone, worldwide, they provide a foundation on which all of us can build.

28. While XML provides an exciting new vehicle for the establishment and maintenance of standards-based directories, it is likely as those directories evolve that they will come to embody much, if not all of the intelligence previously embodied in previous directory standardization initiatives, e.g., X.500. The NIH (not invented here) syndrome seems to be a powerful motivator for those who bear no personal liability for the cost of reinventing the wheel. At the same time, it should be noted that the cost of the learning, implementation, and support curves imposed by complex standards is very high. The KISS (keep it simple, stupid!) principle is generally more effective, unless one happens to be fortunate enough to wield monopoly power to enforce one's will upon the marketplace.

29. For purposes of records management, "classes" of documents are called records series. Agencies are free to propose to classify their records series by whatever schema may be logical to support their business processes, including the need for audits and historical reference. However, they are required to consult with their stakeholders and obtain the approval of the National Archives and Records Administration (NARA) for the schedule on which they plan to destroy any record. It is illegal to destroy any record except under a schedule approved by NARA.

30. The source of this definition of DTD, as well as other information in this section, is Appendix B of Open Buying on the Internet and Extensible Markup Language: Recommendations on Adoption by the Federal Government, by Stephen Luster, Theresa Yee, Mark Crawford, Robert Parker, Christo Andonyadis, and Daniel Drake of the Logistics Management Institute (LMI), under contract to the General Services Administration (GSA). LMI's report is available at http://ec.fed.gov/OpenBuyingXML.pdf.

31. xml.org provides a registry of XML schemas at http://xml.org/xmlorg_registry/index.shtml

32. Voluntary consensus standards may not develop or may take longer than necessary to develop in the private sector because vendors have every incentive to resist if they believe they can make more money by selling proprietary products. That is especially true if any vendors believe it may be able to command a monopoly ("winner take all") position in the marketplace. The best, indeed, the only hope of overcoming this dynamic is for customers to "just say no" to proprietary stovepipe applications. If we the people aren't smart enough to look out for our own interests, we certainly can't expect others to do so for us, not even Uncle Sam.

Companies have a particular incentive to resist the establishment of standards-based person directory services because they will eliminate the need for companies to maintain their own directories of personally identifiable information, and thereby their ability to attempt to "own" their customers under the guise of "customer loyalty." In a truly free market, repeat business is engendered by delivering best values in products and services at each "moment of truth" - rather than by commanding customer information so as to be able to out-promote one's competitors.

(Jan Carlzon, former president and CEO of Scandinavian Airlines, turned the his company into one of the world's most profitable airlines by focusing attention on the customer - rather than the product. Carlzon considers each and every transaction with each and every customer to be a moment of truth.)

However, enterprises such as the Sun/Netscape Alliance are now pushing "customer intimacy." Apparently, customer "loyalty" is no longer good enough for those who fear that their products and services cannot stand on their own merits in the marketplace. A single "host," however beneficent, simply cannot be "intimate" with thousands, hundreds, or even dozens of individual human beings. The thought is not merely promiscuous, it is a contradiction in terms. Moreover, it demeans and sacrifices true intimacy at the altar of pseudo-realistic techno-speak.

The value proposition for directory services might be summed up in the following question: Are we "the people of the United States" or are we "the people of the corporations with which we do business"? For that matter, are we "the people of the organizations for which we work"? Are we "products" of those organizations? Or, rather, shouldn't all organizations - especially government agencies - be products of the people? That can occur only to the degree that we the people are freed from the bondage that is implicit in the ownership of our personally identifiable information by myriad organizations which, by the laws of economics as well as nature, must have their own self-interests at heart. We can only be freed from such virtual slavery by a standards-based approach to person directory services.

To twist Martin Luther King's famous admonition, the truth is: "standards will set us free." Otherwise, at the moments of truth, organizations will not, indeed they cannot be true to a vow of customer focus. They simply lack the means, if not also the will, to do so.

33. OMB Circular A-119 requires agencies to participate in the development, implementation, and use of voluntary consensus standards, a term that is defined in the circular, at http://www.whitehouse.gov/OMB/circulars/a119/a119.html

34. The Wf-XML standard is available at http://www.aiim.org/wfmc/standards/ Joseph J. Rogowski of the JCALS Advanced Concepts Office is a voting member of the WfMC. (See description of JCALS in end note 36.)

35. In large measure, information overload and even "work" overload is self-inflicted - because we have unwittingly bought into the "push" paradigm instead of insisting on the "pull" paradigm. Implicit in the push paradigm is the thought that someone else, or even a machine, knows better than we do how we ought to be spending our time. XML can be used to facilitate either paradigm. However, wise human beings will take it upon themselves to determine when to retrieve the exact information they need or desire, and what to do with it, rather than relying upon other humans or machines to select both the information as well as its time of delivery. Likewise, wise organizations will trust their people to make such decisions, rather than trying to arbitrarily impose information and work tasks based upon outmoded and often irrelevant hierarchies.

36. In a May 26, 2000, E-mail exchange, Michael Rossi of Computer Sciences Corporation (CSC) summarized a WfMC project underway in the Department of Defense (DoD):

JCALS is a DoD project crossing all military services, and managed by the Department of the Army. It provides a technical computing infrastructure (hardware, software, deployment and maintenance services) in support of the DoD's logistics processes. As part of this infrastructure, the system provides a GOTS Workflow Management capability. This capability was used to prototype an early version of Wf-XML, and is now being upgraded to support the official 1.0 version. We have one implementation of the prototype currently being deployed to the Defense Logistics Agency (DLA), and are on schedule to utilize the upgraded function in an Army PDM integration project.
37. Adherents to the increasingly discredited bureaucratic, hierarchical style of management may feel that it is the inherent right of "supervisors" to use automated tools to "flow" work to their subordinates. Indeed, there are even subordinates who may view workflow automation as a means of highlighting the inadequacies of their "superiors," e.g., by automatically documenting their failure to move work products on schedule. However, in neither case is workflow automation an appropriate tool, nor is it likely to be an effective substitute for personal responsibility or for effective leadership in establishing priorities and delegating authority within organizations, much less across organizations. In short, workflow automation is best used to flow well-organized electrons (i.e., E-records in well-formed XML documents) rather than for trying to move human beings as organized (regimented) automatons.

38. One definition of "document" is "data in context." Context is essential for meaning. Data without context is meaningless, to machines as well as to human beings.

39. In light of the importance of records management in the legal community, perhaps it is no surprise that Legal XML, founded in November 1998, was among the first of the affinity groups to recognize the potential of XML. Their home page is at http://www.legalxml.org/.

Already a company called E-Filing.com has established a Web site employing Legal XML to enable users (for a fee of $2.00) to fill out and transmit forms. In an E-mail message dated June 1, 2000, Charles W. Beachboard, Director, e-Business Development, made the following offer:

I am offering your government agency free, off the shelf technology to allow citizens across America as well as your local citizens, the convenience of submitting filled forms/permit applications with applicable fees to your departments electronically with B2 level security over the Internet using the exact forms currently used on a day-to-day basis by those departments. This system cost ZERO tax dollars allowing you the benefit of utilizing IT dollars in other areas and not using your servers to store the fill-able .pdf forms. The data from the electronic form is in Legal XML and compatible with any database allowing routing of information among need-to-know local government agencies and agencies across America if necessary.
E-Filing.com's home page is at http://E-filing.com/. Their Gov2U forms site is at http://www.xforms.org/Gov2U/ They call themselves "America's Virtual City Hall," at http://XForms.org/

40. ISO/IEC 11179 is an international standard for the specification of data elements, including metadata elements.

41. Records management requirements are routinely ignored in planning and designing IT projects and systems even though 36 CFR 1234.10(d), (h), and (m) and 1234.20(a) require agency heads to:

42. Paraphrased from "XML and Accessibility" by Daniel Dardailler, Project Manager, Web Accessibility Initiative, at http://www.csun.edu/cod/conf2000/proceedings/0103Dardailler.html

43. The Center for IT Accommodation (CITA) at GSA has high hopes for the application of XML to facilitate accessibility. CITA's home page is at: http://www.itpolicy.gsa.gov/cita/

44. FIRM's E-records Standards Committee is working on a proposed set of metadata for establishment as a standard or best practice for the classification of records at the series (logical collections) level, Governmentwide. Several of the previously proposed metadata sets are identified at: http://mysite.verizon.net/ambur/RMmetadata.htm

The specific set of metadata elements that are adopted for any particular purpose or set of records is not so important as the need to provide some amount of metadata. No set will ever be perfect or fully comprehensive. If 80 percent of the value can be achieved by using 20 percent of the possible elements, that may not be a bad value proposition. Again, however, the flexibility and extensibility of XML are key. Even if we only start off with a 20-percent solution, we'll be establishing a sound foundation on which to build.

45. Lack of immediate and direct benefit to E-record authors is also a problem, since they may lack appropriate incentives to provide the necessary metadata. However, that problem should be addressed on at least three fronts: a) tools should be provided to automate as much of the process as possible and to make it as easy as possible for people to provide metadata that cannot be captured automatically; b) authors should be educated on the longer-term benefits to themselves and their organizations if their E-records are more readily available and usable; and c) organizational leaders should take it upon themselves to recognize and promote the strategic value of the corporate knowledge embodied in the organization's E-records.

Moreover, it would be unfair to assume that authors - particularly those who have chosen public service as a career - are unwilling to devote any amount of time to classifying their E-records without first giving them the best available tools with which to do so. At the very least, there will be some authors who are not only willing but anxious to make at least some of their E-records readily available ... and that set of records may be a very good place to start.

One thing that should be understood is that, in the context of IT systems and E-records, "management" and "metadata" are virtually synonymous: Little or no metadata quite literally mean little or no management. As with any type of management, the objective is not to micro-manage or over-control the process but to apply the appropriate degree of management.

46. Among a list of issues on which the Senate Governmental Affairs Committee is seeking public input via the Web is whether a "Government Extensible Markup Language (GXML)" should be developed to make government information easier to find. The site at which comments are invited is: http://cct.georgetown.edu/development/eGov/

47. The home page for the Extensible Business Reporting Language is at http://www.xbrl.org/

48. Eric Cohen is maintaining an XML for Accountants page at http://www.computercpa.com/xml2.html

49. The source for these definitions is LMI's Open Buying on the Internet report for GSA, page B-6.

50. The W3C's RDF FAQ are at http://www.w3.org/RDF/FAQ

51. Yet another creative use of RDF might be to establish a directory of the expertise and skills possessed by individuals, which is what many people mean when they talk about "knowledge management." In other words, RDF might be used to make explicit the capabilities that each of us possesses, as individuals and as members of organizations.

52. xml.com provides links to XSLT resources at http://www.xml.com/pub/Guide/XSLT

53. Exhibits 52, 53 and 300B are required under OMB Circular A-11, which sets forth the process for annual requests for funding in the President's budget. The Information Technology Investment Portfolio System (ITIPS) is being implemented to support the process. ITIPS itself should support XML and it should also provide for the consideration of the XML standard in any IT project that is documented within ITIPS.

54. The ebXML home page is at http://ebxml.org/. UN/CEFACT's home page is at http://www.unece.org/cefact/. Information on OASIS is at http://www.oasis-open.org/

55. Shawn McCarthy warned readers of Government Computer News (GCN) to "Get ready for the next big thing: XML" and he noted that "Extensible Markup Language's tag sets will give Web developers greater access to databases." (January 24, 2000, available at http://www.gcn.com/vol19_no2/reviews/1174-1.html.) However, he subsequently suggested that seven "political" problems cannot be solved by XML. (See "XML can solve data distribution problems but not political ones," GCN, June 5, 2000, available at http://www.gcn.com/vol19_no14/com/2100-1.html.)

56. WordPerfect 9 incorporates XML features. The U.S. Patent and Trademark Office is piloting its usage with a group of law firms who file patent applications. A plug-in is available for Microsoft Word.

Corel has been a leader not only in XML but also its older, more complex "parent," SGML. A white paper on Corel's implementation of XML in WordPerfect 9 is at http://www.corel.com/Technical_Marketing_Documentation/WordPerfect_Office_2000/XML/wh-wp-09-win-xml-int.pdf.pdf.

Among the many vendors who are rushing XML applications to the market is a TENdotZero, a company located in Borehamwood, Herfordshire, UK. The company has released a new product suite designed to give small and medium-sized businesses the ability to handle EDI transactions securely over the Internet. The company says the software is fully scalable, works across Internet protocols, and integrates into existing systems, unlike traditional EDI. The company's EdiBus product is an XML/EDI suite of applications that includes: EdiMail, an XML e-mail client; EdiPad, an XML letter-authoring application; EdiOrder, an ordering application which can connect to any database; and EdiControl, a group of administrative controls to help manage the suite. Finally, the company notes that all transactions can be carried out on the Internet and thus can be hosted by an application service provider. The company is looking for BETA testers for its new XML word processor. TENdotZero's home page is at http://www.tendotzero.co.uk/

Another vendor that has literally staked its name on XML is thinkXML, whose home page is at http://thinkXML.com/index.asp. They claim their "e-commerce tools rapidly create complex, highly scalable and maintainable B2B systems, seamlessly integrating diverse front ends with diverse backends using XML technology." And they say their "Transaction Workflow Server II will provide the application richness needed by sophisticated transactions along with the intuitiveness required by casual users."

Microsoft's XML Developer Center is at http://msdn.microsoft.com/xml/default.asp. Their XML Developer's Guide is at http://msdn.microsoft.com/library/default.asp?URL=/library/psdk/xmlsdk/xmlp91b9.htm

Apple says its WebObjects platform provides full support for XML, at http://www.apple.com/webobjects/techspecs.html . They also tout XML support in an article by Ed Siebel entitled "E-Filing and the Macintosh," at http://www.apple.com/smallbusiness/legal/efiling/index3.html . Media Design in Progress says Emile is the first XML editor for the Mac, at http://www.in-progress.com/?id=r8kum

An XML search of Linux.com turns up a number of references, at http://linux.com/search.phtml?search=XML, including a notice concerning Allaire's support for XML in Cold Fusion, at http://linux.com/newsitem.phtml?sid=1&aid=4762

Uche Ogbuji's two-part article on "Practical XML with Linux" begins at http://www.linuxworld.com/linuxworld/lw-1999-09/lw-09-xml2.html. The Linux Documentation Project (LDP) is referenced in an article entitled "Simplified DocBk XML on the Web: Introduction to DocBook" by Jonathan Eisenzopf, at http://linuxtoday.com/stories/18636.html

IBM's XML home page is at http://www.ibm.com/developer/xml/

IBM/Lotus' XML home page is at http://www.lotus.com/developers/devbase.nsf/homedata/xml

An article on Novell's DirXML is at http://www.novell.com/lead_stories/1999/jul12/

A paper entitled "XML, Java, and the future of the Web," by Jon Bosak of Sun Microsystems is at http://metalab.unc.edu/pub/sun-info/standards/xml/why/xmlapps.htm. Other Sun white papers on XML are at http://www.sun.com/software/whitepapers.html#xml. See especially "Portable Data/Portable Code: XML & JAVATM Technologies," by J.P. Morgenthal, Director of Research, NC.Focus, at http://java.sun.com/xml/ncfocus.html

A set of links related to XML, Netscape, AOL, and Mozilla are at http://www.xml.com/pub/Guide/Netscape_(AOL)

A paper entitled "Using XML in Oracle Database Applications" by Brad Wait of Oracle Corporation is at http://technet.oracle.com/tech/xml/info/htdocs/otnwp/about_xml.htm

Sybase's technical white paper entitled "Using XML with the Sybase Adaptive Server SQL Databases" is at http://www.sybase.com/products/databaseservers/ase/whitepapers/L01064.pdf#xml

Informix's XML home page is at http://www.informix.com/xml/

Software AG addresses the question, "What's the big deal with XML?" at http://www.softwareAG.com/corporat/news/May2000/Tamino_open_data.htm

The Apache XML Project is documented at http://xml.apache.org/

Robin Cover maintains an extensive list of publicly available XML software, at http://www.oasis-open.org/cover/publicSW.html

57. Even as the inadequacies of HTML have become apparent, the proponents of SGML argue that the weaknesses of XML will soon catch up with us too. (Corel's WordPerfect already supports SGML, as well as XML and PDF.)

58. In order to be authoritative, rapidly disseminated, and freely subjected to continuous improvement, knowledge must be made explicit. That is, it must be documented records that persist for some time in readily accessible format. The XML standard makes explicit a body of knowledge that fosters the codification and sharing of other nuggets of knowledge. Thus, it enables 6th generation knowledge management.

59. OASIS is maintaining a catalog of XML-related initiatives at http://www.xml.org/. The mission of the site at http://www.xml.com/ is to help us learn how this new Internet technology can solve real-world problems in information management and electronic commerce. The http://xml.net/ domain has also been reserved and is under development. The Development Exchange's XML Zone at http://www.xml-zone.com/ seems to have some good information.  DevX offers an XML Quick Start at http://www.projectcool.com/developer/xmlz/index.html.  The W3C provides an executive overview of XML activity, at http://www.w3.org/XML/Activity, in addition to the XML home page at http://www.w3.org/XML/

The http://xml.gov/ domain has been reserved for use by the CIO Council's XML Working Group. Ideally, it would become a portal site for information on and demonstration of government-related XML activities at all levels of government - Federal, State, and local - as well as internationally.