The Open API & Beyond Builder Techday event organised by CNet Networks Japan and Six Apart was held at Bellsalle Yaesu, near Tokyo Station on the 28th of January.
The event was oversubscribed, and I missed out on the draw, but Gen Kanai from Mozilla came to my rescue and helped me get a pass to join the event. Han Kessels from Tech Talk Japan was also there, and we both managed to charm our way into passes for the BOF and drinks session in the evening (many thanks to the tall lady from CNet!)
He opened talking about Twitter and Facebook. They are successful merely by the fact that their APIs are documented. Same for LivedoorAuth, upon which many developers in Japan have created services, and has now been made compatible with OpenID.
He then contrasted this with Mixi, which has no API, yet there are open Perl tools available to scrape data from the site. The message here is that even if you don't intend to open up your site, users will find a way to get their data out anyway.
Open APIs create combined value, drive traffic and increase exposure. Nothing revolutionary here, but he illustrated the idea that opening up your data to third parties is a great way to generate more traffic on his own blog/aggregation site. There he takes data from various APIs, including Flickr. Images from Flickr are linked back to Flickr, so it's a win win situation. While this may work for the example given, sites that have grown on the basis of keeping data locked in may have trouble dealing with this.
Which was illustrated quite nicely with a quote from Anil Dash, and by the evolving 'net history.
- Email Messages were orignally stored on one machine, but as networks evolved, they became SMTP.
- IM wars are slowly giving way to Jabber/XMPP based messaging.
- On Arpanet, all host names were stored in the hosts file, but that became the DNS
- Similarly, centralised web systems are transitioning to open platforms.
This natural progression to openness is not easy, but we'll get there in the end.