Wikipublisher automates typesetting so authors don’t have to worry about it, is that right? Well, not quite.
Copyfitting is the process of adjusting type to fit the space available on the printed page. Web authors typically don’t need to think about this — web pages are so malleable that attempts to control the fit are generally over-ridden by individual readers’ browser settings. Wikipublisher takes advantage of LATEX’s typesetting capabilities so that the same is almost true of the print output produced. If the web page is OK, chances are the print page will be too. Up to a point…
The first issue facing the editor responsible for final copyfitting is the expected print metadata settings. If a reader changes the paper size, paragraph separator (space or indent) or fontset, the fit also changes. If a page is printed on its own or as part of several lists, the pagnation is different in each context. So there is no right answer.
Experience to date shows the following are the most common copyfitting problems:
The solution is simple: re-word one or more paragraphs. Some people may say that an author should not need to re-write something to compensate for the imperfections of a typesetting program. However, most writers take the view that no paragraph is perfect and the copyfitting process is an opportunity to improve the text. Following The Elements of Style (Strunk and White), this writer’s strategy is to look for, and omit, needless words.
LATEX uses a complex optimization algorithm to decide the best layout for a page. If one of the above problems arises, we can be confident that all the alternatives are worse. The parameters (“penalties”) controlling the algorithm can be tuned, and it is possible that different values may produce better results than the present settings. Any advice on this topic would be welcome.