To start, you need an ISBN from a paper catalogue, publisher's mailing or publisher's web site.
Through this interface, you can add new books, change any of the details of existing books or delete an individual book. Changes made in the database are not immediately reflected on the web site: Martin has to log into the server and refresh the site.
See the separate article about Book Catalogue Quality Control
You can start by typing in either an ISBN10 or ISBN13. 13-digit ISBNs will be automatically converted to 10-digit.
When you type in an ISBN, the page will look in the existing database. If it finds a book, it will bring up its details, including subject classifications. If the book has been previously catalogued and then deleted, it will tell you this: this usually means that the book has been superseded by a more recent edition.
If it does not already know about the book, then it looks it up on Amazon. If there is no information for that ISBN on Amazon, you get a blank form.
If a book is very new, it might not be listed in Amazon, in which case you'd have to put author, publisher and so on in manually (preferably by copy-and-paste from the publisher's web site).
The script knows how much the books cost, and will automatically decide if a book counts as "Library" or not.
The "Annotations" at the bottom of the form determine which subject sections of the guide the book appears in. Each book must have at least one annotation (or specific country/economy). If you want to add or change a book's subject annotation, change the "subject", then the "subsection", then press "Put this in the database" and the record is updated. The order in which the annotations appear on the form is irrelevant.
The data we get from Amazon do not include textbook companion sites or publisher's links. Hence it is a good idea to periodically look at the sites of major publishers, which usually have dedicated sections for companion sites. If you click on a publisher name on the right of the screen, the script will "guess" the URL for publisher information on the book.
The Amazon listing does not include an edition number. It's a good idea to type this in if it applies (e.g. if it's mentioned in a paper catalogue), so that if we find two different sets of details for the same textbook, we know which is more recent. An edition could also be a phrase, such as "European edition".
Bulk deletion is possible through the Access interface (run by Martin) and we hope to develop a web-based bulk deletion system. At the moment you can delete individual books using a button on the bottom right of the cataloguing form. This button copies the book data to a hidden area and removes it from the live area of the database.
Copying the book data back if there is a mistake requires directly connecting to the database server; in other words it's not simple. Basically, only delete a book if you're sure we never want it to appear on the site again.
At the moment, the system uses 10-digit ISBNs. It could change over to ISBN13, but this requires some programming work. While ISBN10s are ubiquitous, we'll keep using them.
COPAC is a service which brings together some of the UK's main academic research libraries, mostly from prestigious universities. Each item in the book catalogue has a link to a COPAC search for its ISBN, which we're told is the most efficient way to find the book on the COPAC site.