Ideas for future development

Ideas and suggestions for future development of Open Library. If you can, be specific about what you would like to see.

Easy add of new books

It should be easier to add a new book, in a way that doesn't require the user to key in all of the bibliographic data. If you have used LibraryThing you know that adding a book merely requires providing either an ISBN or some information like the title of the book. This data is then used to search a selection of catalogs (Amazon, Library of Congress, some other library catalogs) and pull up data that matches the search terms. For OL, when a user selects an item from the list, it would fill in that bibliographic information in a new edit screen.

Keeping OL up to date

For a while, OL had a subscription to LC Books All, which provided weekly updates to LC cataloging. It also ran scans of Amazon for new material. At the moment, none of these updates appear to be happening.

LC's catalog provides identifiers for books that are sequential, with the format:
http://lccn.loc.gov/yyyy999999

MARCXML is available by appending /marcxml to the URL e.g.
http://lccn.loc.gov/2013000001/marcxml

Other formats available using the same scheme include MODS, MADS, and DC.

LC has APIs that allow the download of MARCXML records. Presumably we can determine the missing record identifiers and download those that are missing, as well as run periodic queries to keep OL up to date.

Data Cleanup

Data quality (or lack thereof) is one of the biggest problems that we have right now. Below is a partial list of some of the known problems which need cleaning up.

• encoding - both html entity encodings & bad character encodings (e.g. UTF-8 interpreted as ASCII)
  
• name order - many author records still have inverted name order
  
• normalize ISBNs - a small percentage of ISBNs are not in a common format (eg ISBN-10 or ISBN-13). See link text for list.
  
• [from old catalog] strings sprinkled throughout dates and other parts of records
  
• duplicates - lots and lots of duplicate author and edition records
  
• empty author records imported from LC catalog or IA without any works attached
  
• square brackets in names and dates - sometimes one each place where a name/date string was apparently split apart
  
• Internet Archive imported books with only edition records, not linked to any work record (and thus no author link)
  
• Bots for attacking spam?

Improve EDIT pages

• Add "other_titles" to edition edit page (openlibrary / openlibrary / plugins / openlibrary / types / edition.type)
  
• Add some mouse-over or click-on help bubbles to explain more about some of the fields
  
• Make all items in Open Library editable (including IA-imported items)
  
• Make some illustrated help files to explain non-intuitive aspects of the EDIT feature
  
• Make Librarian mode the default. I don't know why we hide these fields in the first place.

User-facing features

• Searchable lists/users
  
• Fuller-featured profile pages & possibly social options
  
• Early book returns for people using non-ADE checkout options
  
• Automate the signup process more so users can fix a typoed email address & resend verification email to themselves
  
• Remove non-functional features (list export, merge editions, delete account)