 |  |
| |
Creating a data format is often only the first step in making it useful. If an XML vocabulary is only used for a particular process inside a software application, there may not be much reason to publish information about how it works, except for future developers who may work on that application. If, on the other hand, the data format is intended for widespread use by people or organizations who may not normally interact with each other beyond the exchange of messages, then it probably makes sense to provide much more support for the format.
There are a variety of different kinds of information about a data format that are frequently worth sharing:
Human-readable documentation, perhaps even in a variety of languages
Schemas and DTDs formally defining the structures and content
Stylesheets and transformations for presenting the data or converting it from one format to another
Code for processing the data, perhaps even in a variety of languages or environments
The first two approaches--human-readable documentation and schemas--are typically the foundations. Formal definitions and rough understandings of what goes where often work for formats that are used by individual programmers or small groups, but sharing formats widely often requires further explanation. Stylesheets and code are additional options that may simplify adoption for developers.
The appropriate level of publicity for an XML vocabulary can vary widely, from no publicity at all to publishing a RDDL document or a support site, to registering the format in one of the XML application registries, or to creating a working group at some kind of standards body or consortium.
|  | |
| 15.2. Developing Data-Oriented XML Formats |  | 16. XML Schemas |
Copyright © 2002 O'Reilly & Associates. All rights reserved.