Rambling About APIs
I've been thinking about turning my Amazon scraping scripts into an XML API to their book information (I call these SCRAPI
s), but it could never be as reliable as Amazon offering their own API. Plus I'd have to keep up with their page design changes. It's fun to think about rogue APIs to web sites, though.
I wonder when/if eBay will open up an interface to developers. I bet there are thousands of home brewed scripts out there to scrape eBay for auction information. If they had an API, I could see specialized auction sites popping up with their own design...and eBay would always get a cut of the transactions. Perhaps some sites would like to offer auctions, but they don't have the resources to develop their own software, and they don't want to send people to eBay. Yet they have an audience that would be willing to participate. The value in offering an API is extending services to places they can't go today. And in the ideas generated outside of the company...those most interested in the service become the development team. It's sort of like widespread prototyping.
This can be scary for companies. Michael Schrage said, "Prototype-driven innovation ends up promoting a radical deconstruction of existing organizational charts..." In the sense that teams form around prototypes, rather than teams being put together specifically to build them. I think on the outside, through an API, this behavior works to the company's advantage. It attracts people to their technology, it brings people in with fresh ideas, and it discourages politics that have formed around ideas. If you think about an API as widespread prototyping, this interview
with Michael Schrage about "shared spaces" has some great ideas; even though he's mostly talking about working within