The FAQ says "PDF is a hard-to-read format". In fact, it's not outrageously difficult to read. And it is a properly documented standard. The difficulty is that people often imagine that they will be able to "convert" a PDF file to some other format that has a different purpose.
A PDF file contains a document for display and/or printing. It contains instructions (very much like compiled PostScript) to draw lines, shade areas, write text etc, at various places on the page. It does not contain much, if any, metadata about the relationships and purposes of these lines, areas and text. In this respect, it is very different to things like HTML,
Word documents,
RTF etc.
As an example, say you have a PDF file that, when displayed, shows a table of values. Nothing in the PDF file says it's a table. It's just a load of lines and text in various places. Therefore, it is nearly impossible for a general program to identify the PDF as a table and convert it into, say, an OpenOffice Document containing a table.
(To bartenders: I've written a number of similar replies recently. Any chance of improving the FAQ entry? I work for a company one of whose main businesses is PDF, so I'm sure I can get a nice concise entry for you.)
