O’Reilly Radar posted yesterday about OpenLibrary. So right away I thought of LibraryThing and find out that Tim’s already involved. This seems like a great idea for several reasons. First there’s no proprietary control of the data as with Google Books. Second is that there’s a open wiki, it seems mostly for metadata but other comments too.
I hope they add useful APIs to the project so that users can build on the data to lump and split it the way that is most useful for them. I don’t think you can have an “open” project of this type and scale without an API. And using an image viewer that restricts viewing to certain browsers doesn’t seem very open (Safari seems to work in spite of the warning).
I’d like to see the wiki idea extended so that users can comment on a specific section in the work. With comments attributed to a particular user, a good API would let me see the comments of the users which I respect. Use an OpenID based attribution system so that I can combine it with the reputation system that works best for my goals. Suppose you were taking a literature class, the teacher gives an assignment to read and comment on a chapter in Tom Sawyer. The book would appear on the left side of the screen and the comments on the right side. Your viewer using the API would filter out the comments from non-class members which could lead to an interesting discussion led by the teacher. Seems perfect for distance learning.