Information Modeling with XML
1.1 Introduction
When XML first came into use, it was seen primarily as a data interchange standard. Since then it has come to be used for more and more thingseven serving as the core for development and deployment platforms such as Microsoft's .NET. Increasingly, XML has become the means to model components of information systems, and those components automatically construct themselves around what has been expressed in XML. This represents the real potential of XMLthe ability to model the behavior of an entire application in XML once, instead of repeatedly in different ways for each component of an application program.
As long as XML was used as a container for data managed by legacy systems, it was sufficient to consider only syntax when building documents. Now that XML is being used to do more than simply express data, it is important to consider grammar and style as well. Obviously, proper syntax is necessary for parsers to be able to accept XML documents at all. Good grammar insures that once XML information has been assimilated, it can be effectively interpreted without an inordinate need for specific (and redundant) domain knowledge on the part of application programs. Good style insures good application performance, especially when it comes to storing, retrieving, and managing information.
Proper XML syntax is well understood and documented, so that topic will not be discussed here. This chapter does not discuss how to build XML schemas or DTDs, as they are also well documented elsewhere. This chapter is intended as a practical guide to achieving good grammar and style when modeling information in XMLwhich translates to building flexible applications that perform well with minimal programming effort. Grammar is often regarded as being either right or wrong. True, there are "wrong" grammatical practices; but past that, there is good grammar and bad grammarand everything in between. Arguably, there is no such thing as wrong style, only a continuum between the good and the bad.