Blog

I’ve thought for some time that it would be a great project to create some sort of web-oriented Delicious Library. So when I discovered LibraryThing through a stray Google search, my first response was delight that someone (Tim Spalding) had beat me to it. LibraryThing allows you to create an online catalogue of your books collection. It spares you the tedium of typing out loads of bibliographic information by matching title words or an ISBN (for example) with Amazon’s databases and the Library of Congress catalogue. More excitingly, LibraryThing is designed to bring people together through books. If you make your catalogue public, then other users can discover that you own a lot of the same books, read your profile, browse your collection, and peruse your comments. (You can see the books I’ve catalogued here.) It’s a shame users can’t rate books, however, as ratings have greater currency when you know a bit about the judge (which is the problem with Amazon ratings). The predictable enumeration of the most catalogued books looks like it’s going to be as dull and useless as Technorati’s unchanging list of popular blogs (yes, Harry Potter and the Half-Blood Prince is at the top). It would be much more interesting to have a list of books chosen at random, along with their cataloguers’ tags, comments, and reviews. LibraryThing is free to join, but if you want to catalogue more than 200 books, you have to pay a one-off fee of $10 (may rise) for lifetime membership.

LibraryThing is currently still in beta, but has already been reviewed by Andrew Brown (at helmintholog and later in The Guardian) and by graphic designer Stefan Hayden, who notes some usability problems. A Young Librarian notes that librarians would find LibraryThing incredibly useful for reader advisory services, while Steven M. Cohen wishes it could interface with local libraries like Elf, which sends you emails when books are due, overdue, or ready for pickup.

On his blog, Brown praises LibraryThing for prioritising accurate cataloguing over good looks:

There’s a huge emphasis on [displaying] the covers of books [among LibraryThing’s competitors], and all the lookups are done from the Amazon catalogue, rather than the LoC [Library of Congress].

Librarything [sic] uses Amazon as a fallback and (at my request) will also use the European amazons [sic]. But Spalding would rather use the British library, which has to be the right thing to do. From a cataloguing point of view, I really don’t care what the cover of the book looks like. The only useful service would be a picture of the spine, and no one provides that.

Personally, I see no reason not to demand both aesthetics and accuracy. Ignoring the clever AJAX, the look of the LibraryThing site is workaday. If you have a mixture of books with and without cover pictures from Amazon, your listing looks terribly uneven. Ideally, LibraryThing should fake missing covers and/or allow users to upload their own images of covers. A simulated bookcase view, like Delicious Library offers, would also be fun. Brown’s point about spines is well taken, however, and if some web service like Amazon would start providing images and dimensions of the spine then it would certainly be possible to scale them to fit virtual bookshelves like those designed by iBookshelf (here’s a good screenshot).

Mind you, pace Brown, LibraryThing isn’t all that great for cataloguing either. Whereas KDE-based cataloguing software Tellico has access to several library catalogues (and you can add new ones), LibraryThing is currently only searching one. Whereas Tellico comes with more bibliographic fields than you want (and you can add new ones, see this screenshot), LibraryThing doesn’t yet offer anywhere near the same level of detail. Entries are limited to the following fields:

• title;
• author;
• date;
• Library of Congress call number;
• ISBN;
• Dewey number;
• publication (information);
• summary line;
• tags;
• comments; and
• review or review URL.

Spalding is planning to add user-defined fields. This should make LibraryThing ideal for collectors, who will want to add information about the condition of their books. Obsessive bibliographers should note, however, that the title field is too short for long titles and the publication field is too short for detailed information. Presumably, user-defined fields for editor, translator, and so on will not pull in information from the MARC library catalogues. If Spalding were to define such fields for all users, then LibraryThing could become a goldmine of readily accessible bibliographic information in its own right. In any case, there needs to be an FAQ on details such as how to enter special characters and multiple authors.

Spalding’s design may have some growing pains, but the crucial social aspects of the site seem well implemented. Unless a competitor starts grabbing more bibliographic data from more libraries, I’m likely to stick with LibraryThing.