Why does CML stress the component-based approach?
In paper-based publication the medium and message are inextricably linked and can normally only be processed by humans. The development of electronic scientific publishing has often involved conventional “paper” documents with associated electronic data. Thus the electronic deposition of crystal structures and macromolecular sequences is now a routine part of publication. In general, however, the document part of the paper, though often carried in electronic form, is conceptually based on a conventional paper-based document structure. A major feature of this approach is that form and content are mixed. For documents to be reliably machine-processable the paper image is not sufficient. It is necessary to identify the various components of a document both in their intrinsic nature and their role in the document structure. This process is termed markup and has been adopted by many publishers through the Standard Generalised Markup Language (SGML, ISO:8879).
