Top-end library automation software
is remarkably complicated. It generally includes these functions:
ACQUISITIONS: Ordering books from different vendors via several different electronic interchange standards, managing purchase orders and shipments, and automatically notifying vendors of missing items. Ideally, the process of selecting books to buy would be made easier too, with an "Amazon-like" service that provides rich information about the titles, authors, etc.
CATALOGING: Creating data records in the MARC format, a flat file with numbered fields, some of which are repeatable, all of which can have subfields, some of which are repeatable. Verifying that the records are properly formatted, and also helping the cataloger to use the standard library vocabulary. That controlled vocabulary is sometimes unintentionally hilarious:
When delivering speeches Berman would often rely on a light bulb as a prop, which he would hold aloft and ask his audience to identify. Light bulb, you say? Not to the Library of Congress, he'd explain, for whom the correct answer is "electric lamp, incandescent." Such convoluted labeling, Berman believed, did little to promote the nation's 16,000-plus public libraries as storehouses of knowledge designed for citizens who know a light bulb when they see one.
CIRCULATION: Checking books in and out; sending overdue notices to library customers by various methods (by telephone with voice synthesis, by snail mail, and sometimes by email). Keeping track of who's in line for the latest Stephen King book. Optimally filling seven requests at four different branches for an item that's owned by ten different branches. Taking inventory by scanning barcodes of the items actually on the shelf, then identifying which items are missing and which items are shelved in the wrong order... preferably in real time so the book can be reshelved immediately.
INTERLIBRARY LOAN: Finding out who's got (and is willing to lend) a copy of a particular out-of-print book; sending the request; finding out which libraries still owe us books that we lent them...
PUBLIC ACCESS: The various front ends that you, the library customer, can use. Sometimes it's web-based (but see http://suncat.tblc.org/ for a web-based catalog that doesn't work for blind people, WebTV users, etc.), sometimes it's telnet-based (telnet:email@example.com), and sometimes it's an application that runs on a PC. My library system happens to use all three. Ideally, it should be able to use some sort of secure connection so that Little Johnny Snortkiddie doesn't find out who's checking out what books. (Let me detour to talk about confidentiality. It's a basic principle of library service, for very practical reasons. If a message were sent to TheJonesFamily@aol.com saying "Your book 'How to safely leave an abusive husband' is overdue", it could mean someone's life.)
SERIALS: Magazines, newspapers, etc. are a hairy problem. You don't want to catalog every issue of Creative Loafing as a separate item, but you do want to know that such-and-such issues are on microfilm, others are in newsprint, these issues have been sent to the microfilm vendor, and this issue is missing. Then magazines will change their names, change their regular publication pattern (Jan/Feb 2001, Mar/Apr 2001, Special May Edition, where the hell'd that come from?)...
STATISTICS: And, of course, for all of this, we have to have statistics. Libraries live and breathe by statistics. "You can't slash our budget, the taxpayers have been using the library 15% more than last year. You should give us 20% more money for books and bandwidth and staff so we can stay ahead of the curve."
OK, that's Library Software 101. Any questions?