You probably know that XML is descended from SGML, and it thus inherited DTDs, which were then the only means of specifying a document model. That proved adequate for SGML, but its limitations soon became clear in XML. Several alternatives were proposed, but XML Schema seems to be winning. Thus Bradley has delivered a timely exposition. He covers all the features clearly. A brief mention is made of its competitors, Relax NG and Schematron. But they do seem to be fading fast.
The most interesting part of Bradley's text are the chapters on namespaces. Other aspects of Schema are lower level and, while useful, are frankly mundane. By contrast, namespaces are the critical feature of Schema. They let you build on pre-existing schemas that have been published on the web. And you can publish your schemas so that others might benefit. You and those in your field or industry can cooperatively derive a net gain by agreeing and publishing standard definitions. A Network effect.
I assume that you are familiar with HTML. In that, the crucial element are the hyperlinks (the href and src attributes in certain tags). It is these that put the "H" in HTML. Without them, HTML just becomes a limited page markup language. It is that ability to link to arbitrary locations on the Internet which produced the Web. Likewise, in the much heralded Web Services, these exchange data via XML. Which in turn depends on XML Schema to build consistent hyperlinked semantics.
You should read Bradley's chapters on namespaces thoroughly.