As information content on the Web continues to grow and publishers continue to experiment with various combinations of pay and free access, libraries' collections have become more difficult to define. They can change without notice. There are many explanations for the problems and inconsistent reproduction rights. However, it deeply concerns the author that despite the many dollars spent on library databases, a discovery engine, and a link resolver, they still fail. Generalizing is difficult since the failures can occur only for specific titles, during specific time periods, and with certain combinations of bibliographic databases, link resolvers, and publisher's sites. But it is becoming a frequent enough occurrence that the author is starting to check for free copies on the Web sooner than he used to.