XML DTDs Vs XML Schema

DTD vs XML Schema

The DTD provides a basic grammar for defining an XML Document in terms of the metadata that comprise the shape of the document. An XML Schema provides this, plus a detailed way to define what the data can and cannot contain. It provides far more control for the developer over what is legal, and it provides an Object Oriented approach, with all the benefits this entails.

So, if XML Schemas provide an Object Oriented approach to defining an XML document's structure, and if XML Schemas give us the power to define re-useable types such as an ISBN number based on a wide range of pre-defined types, why would we use a DTD? There are in fact several good reasons for using the DTD instead of the schema.

Firstly, and rather an important point, is that XML Schema is a new technology. This means that whilst some XML Parsers support it fully, many still don't. If you use XML to communicate with a legacy system, perhaps it won't support the XML Schema.

Many systems interfaces are already defined as a DTD. They are mature definitions, rich and complex. The effort in re-writing the definition may not be worthwhile.

DTD is also established, and examples of common objects defined in a DTD abound on the Internet -- freely available for re-use. A developer may be able to use these to define a DTD more quickly than they would be able to accomplish a complete re-development of the core elements as a new schema.

Finally, you must also consider the fact that the XML Schema is an XML document. It has an XML Namespace to refer to, and an XML DTD to define it. This is all overhead. When a parser examines the document, it may have to link this all in, interperate the DTD for the Schema, load the namespace, and validate the schema, etc., all before it can parse the actual XML document in question. If you're using XML as a protocol between two systems that are in heavy use, and need a quick response, then this overhead may seriously degrade performance.

Then again, if your system is available for third party developers as a Web service, then the detailed enforcement of the XML Schema may protect your application a lot more effectively from malicious -- or just plain bad -- XML packets. As an example, Muse.net is an interesting technology. They have a publicly-available SOAP API defined with an XML Schema that provides their developers more control over what they receive from the user community.

On the other hand, I was recently involved in designing a system to handle incoming transactions from multiple devices. In order to scale the system, the chosen service that processes requests is a SOAP server. However, the system is completely closed, and a simple DTD on the server is enough to ensure that the packets sent from the clients arrive complete and uncorrupted, without the additional overhead of XML Schema.