I am working on a project to create (in PDF) letter charts for children for alphabet learning. This involves a number of different pictures from different sources (but mostly Wikipedia) and different licenses (but mostly GNU FDL, Creative Commons or public domain). My problem is: how do I properly attribute copyright and licensing of all those different pictures? How do I reconcile all those different licenses, i.e. what license can/should I use for my final work?
Does it make any difference whether we're talking about the source code here (images lying around as project files) or final PDF documents?