WYSIWYM

In a WYSIWYM editor, the user writes the contents in a structured way, marking the content according to its meaning, its significance in the document, and leaves its final appearance up to one or more separate style sheets. In essence, it aims to accurately display the contents being conveyed, rather than the actual formatting associated with it.

For example, in a WYSIWYM document, one would manually mark text as the title of the document, the name of a section, the caption associated with a figure, or the name of an author; this would in turn allow one element, such as section headings, to be rendered as large bold text in one style sheet, or as red center justified text in another, without further manual intervention. More often than not, this requires the semantic structure of the document to be decided in advance before writing it. The editor also needs a system for exporting structured content to generate the document's final format, following the indicated structure.

The main advantage of this system is the total separation of content and presentation: users can structure and write the document once, rather than repeatedly alternating between the two modes of presentation—an approach which comes with its own switch cost. And since the rendering of formatting is left to the export system, this also makes it easier to achieve consistency in design as well.

Document processors

Unlike an unstructured, fixed-presentation document, a document processor rather than a word processor must be used for WYSIWYM. The first document processor which articulated itself through the WYSIWYM term was LyX document processor, although similar concepts can be traced back to earlier publishing systems such as LaTeX, TPS (modeled on pioneering experiments at Xerox PARC) and FrameMaker.

For example, in FrameMaker, a document may be authored (i.e., viewed and edited) using two possible page views, the styling tags (called "formats") and contents of a document reside in "body pages", while structure and presentation are defined by "master pages". Multiple such documents can be linked together and programmed conditionally for a variety of applications. These elements can then be exported to corresponding tags which preserve some or all of their functionality in a markup language such as XML/HTML/CSS and PDF, or directly rendered down for final presentation in anything from PostScript to raw text.

A different approach to the WYSIWYM philosophy is taken by GNU TeXmacs, which combines the on-screen representation of the document structure with an almost-faithful WYSIWYG rendering.