"It is impossible for ideas to compete in the marketplace if no forum for
  their presentation is provided or available." �        �Thomas Mann, 1896

---

XML Trends and eCommerce

Introduction

With the growth of the Internet and the World Wide Web in the sharing and distribution of information, standards have revolved around the presentation of that information, or content, on pages presented in HTML. The standards have been primarily involved with the presentation of graphical information, as is the case of JPG and GIF graphic formats, as well as the structure of text via tag semantics (we are all familiar with the <H2> and <h1> tags).

Although this has been practical, the definition of HTML in terms of new tags and changing technology have been at issue when it comes to the rigid implementations of these tags by browser vendors. Those with the need to disseminate information widely are faced with incompatibilities between how Netscape and Microsoft browsers. This can present a significant roadblock when interpretation of HTML tags is not implemented uniformly between these browsers, or for that matter, between different versions of the same browser.

The implication XML has for electronic commerce is substantial. If we agree to define eCommerce as the use of technology to facilitate business, then the need to deliver an identical result for users of either Netscape or Microsoft browsers becomes a business essential. While the ability to mandate the use of a specific browser in an Intranet or Extranet environment may mitigate this issue, companies and organizations that wish to engage customers over the Internet incur substantial cost to ensure that they deliver reliable transactional results to any customer.

Enter XML. As XML is just a meta-language specifically for the description of markup languages, it is not semantics or a tag set. XML simply provides a facility to define tags and the structural relationships among them. Due to the fact there is not a predefined tag set, there are, consequently, no preconceived semantics. For an XML document, the semantics are defined either by the applications processing them or by stylesheets that can be thought of as XML schema templates.

Let's look at the structure of an XML document.

Like the tags of an HTML document, XML "elements" are composed of start and end tags with data in between. This encapsulated data is known as the value of the element. In the example below, the term Red is the value of the element color:

<color>Red</color>

The color element gives us the ability to indicate (or mark up) the valueRed semantically, which allows us to differentiate bits of data from one another. The same value can be used in other elements as necessary:

<film>Red</film>

The fact that the elements are unique, in terms of their tags, means there is no confusion about the interpretation of the elements by the browser; the first Red could be considered the color of a can of paint, and the second, the title of a motion picture (and a pretty good one).

These elements, as with HTML, may also be nested. Correspondingly, nesting must be maintained without overlapping; and additionally all XML elements must have an end tag, with all attribute values enclosed in quotation marks.

For instance, the following is appropriate:

<XML ID="ABCD">

<music>

<RELEASE ID=1234">

<title>Baltimore</title>

<artist>Nina Simone</artist>

</RELEASE>

</music>

</XML>

The previous example can also be thought of as being a data island, or XML which exists inline within an HTML page. Through this, we can put this code into our page and script against it without having to load the XML through script or the HTML <OBJECT> tag. It can also be referenced in another form, through an SRC attribute on the XML tag:

<XML ID="ABCD" SRC="music.xml"></XML>

When a browser parses an XML data island, an XML document object is created. This object has properties and methods that we can use to code against, or manipulate. The XML parser exposes the object model and allows us to manage XML document nodes. When a data island is parsed, a tree of nodes is created. These nodes can be accessed through script. For example, the second child element node object of the XML document element from the above example <XML ID="ABCD"> would be identified by the following:

XmlDocument.documentElement.childNodes.item(0).childNodes.item(1).text

(The Second child node from the example would return the following...........Nina Simone)

Further, by defining small sets of XML elements, we can create stylesheets, or XSLs. The importance of this is obvious when it comes to filtering or sorting XML data. If we picture the XML as the input to a XSL, we can view the result as well-formed HTML or as output for further XML manipulation.

With XML parsers within popular web browsers becoming more common place, and the corresponding availability of XML enabled application becomes more widely used, we will see a greater level of control and flexibility in transmitting important data. Such scenarios include Internet commerce, where a salesperson will download XML from a company database, and through XSL and client-side browsers, portray the information as an order catalog.

This data will then be augmented by customer purchase information and uploaded to the same database where it would be reflected as ordering confirmation and inventory adjustment criteria, and finally, confirmed to the customer via a browser, where it would be supplemented by secure data encryption elements and presented as order confirmation. The key point of all of this will be the near linear nature of the transaction while being maintained throughout heterogeneous computing platforms, crossing midranges and browsers within UNIX and Windows PCs, midrange hosts, and PDAs.

Add to this the manageability of the data for such mechanisms as search engines, in which XML will allow a greater control of input and true criteria based output through exposed XML properties and methods. For example, imagine querying a search engine for a 1985 White Nissan 300ZX and being presented with a listing of dealerships that have the same product in stock.

With all of this, it is easy to see that while XML is not a vehicle to replace HTML or even SGML (the latter of which is much more complex, but not ideal for web based information deployment), it is a worthy and welcome complement to the current Internet tools needed for serious development.