About a month ago, Jason Kottke elaborated on his assertion that “Facebook is AOL 2.0″, earning a reaction from Jeff Atwood
I feel very strongly that we already have the world’s best public social networking tool right in front of us: it’s called the internet. Public services on the web, such as blogs, twitter, flickr, and so forth, are what we should invest our time in. And because it’s public, we can leverage the immense power of internet search to tie it all– and each other– together.
When I recently mentioned the Good Reads social-book-site, I referred to some thoughts which I didn’t post about rekeying book info, and shortly after I saw Revish, which looks like yet-another social/book site, and the mind boggles further. This is just another facet of the “lock-in” problems discussed by Kottke and Atwood, not to mention many commenters to their blogs. The bogglement has been such that this post has languished unfinished for a couple of weeks, but now Heath has called me out and I feel obliged to try to squeeze it out…
In the vision espoused by Atwood, I’d just have all my own “book information” on my server, and other systems would make use of it. Heath alluded to theFOAF Project as an attempt to do this about people and their interrelationships. Of course, FOAF has not made much mark on internet culture at large, and I think the obstacles it has faced are the same as those facing this hypothetical book ontology. It takes a lot of effort to participate, and there isn’t that much interesting you can do after exerting that effort.
At first blush, the interestingness question is a bootstrap one. If one site pioneered a format, people might well build interesting tools upon it. However, FOAF has been around for a while and the tools haven’t hit yet. There’s a root problem in that “book information” is far less tidy than those two words might suggest. On some sites I indicate books that I want; on some I indicate books that I own; on some, books that I have read or which I want to read. The same sorts of problems apply to social relationships, and in fact, FOAF has borne out the promise of RDF because a number of extensions have been defined. Regardless, the interestingness/effort quotient hasn’t hit the sweet spot.
Perhaps there is some inspiration in the path of simplifying the problem down to common identifiers. For example, all of the half-decent social networking services allow you to import a list of email addresses and leverage the fact that you have an address book into somehow using their site a little more effectively. Similarly, a half-decent book information service should be able to import a list of books (no doubt as ISBNs) and then accelerate through the process of explaining to the service why you care about those books in its terms. I tried exporting my list from LibraryThing and importing it into the Books iRead Facebook application but Books iRead didn’t find any ISBNs. It wasn’t really worth it to me to troubleshoot, because the import only allows me to bulk tag all of the books as either “read”, “reading”, or “wanna read.” But a somewhat more sophisticated interface could load all of my books into a holding pen and then lead me through the steps of integrating them into my profile in the application, including offering to save pending books for me so that I could work on it over time.
Having good common identifiers at least minimizes the lock-in problem. After that, correctly identifying common properties would minimize the tweaking issues. Even so, there’s a long gap between importing and exporting data and providing a master data source that could be used to serve an infinite number of applications. I’d like to see someone lay out the social use cases that would drive any project to tie together public internet services in Atwood’s model.