The PDF solution was bad in many ways; e.g. it didn't even have the ability to colour links, and some internal links were turned into external links. However, at the time when the manual was ready for being included into OpenMPT, PDF was the only format to which I could export the wiki contents to was PDF, since e.g. the DocBook renderer for the export plugin was not available. However, with the help of coda, I created a page scraper to be able to create a collection of exact HTML copies of the articles. Now the manual looks the way it's supposed to look - on screen, anyway. However, it would also be possible now to create e.g. a single big HTML page to print out, if someone really needs that.