During 1995-1996, at the University of Utah, with U.S. National Science Foundation support, we hosted a Special Year in Mathematical Biology . This brought together eighty-six outside researchers in the field (graduate students, post-doctoral fellows, and visiting faculty) from six countries, for visits ranging from a few days to several months, at our Mathematics and Biology departments. At the conclusion of this activity, we set about to prepare a book  to document the research carried out during the Special Year.
I was the consulting technical editor on the book, and its production took almost six months of extremely intense work on the part of the contributing authors, editors, and technical support people. Work began in June, and the day before Thanksgiving in late November, I uploaded the final completed POSTSCRIPT files to the publisher's printer shop. Just five weeks later, the book was available in bookstores.
The book has 573 literature references, every one of which was checked against library catalogs and journal databases. Though tedious, this checking was extremely worthwhile, since it uncovered scores of errors and omissions which were repaired before publication.
Since the book documents forefront research in the relatively young field of mathematical biology, most of its references were to work of the last two decades, and researchers and students who use the book will very likely want to read many of those publications.
With conventional bibliographies, the bibliography is of little use in itself, since one has no easy way to find out where in the text a particular publication was cited, and why, unless, of course, one has an author who is as conscientious as Donald E. Knuth is about indexing cited authors.
I therefore decided that the bibliography in our book would be different: every bibliographic entry would contain a list of page numbers in the book where the publication was cited, and every author of every publication would be indexed. This latter index is important, because good scientific research often requires reading all of the publications of a prominent author in a particular area. In addition, book entries would contain International Standard Book Number (ISBN) values, providing unique identifiers that can be helpful in locating books in libraries, or ordering them through bookstores.
The problem with this objective is that with hundreds of citations, and over a thousand authors in the bibliography, accurate manual preparation of the cross-references and index is humanly impossible. Computers, however, get neither bored, nor tired, so they can accomplish the job, given suitable software. At the start of this project, such software did not exist.