I think OpenSocial has 3 main parts.

1) Write-Once, Run-Many gadgets. Instead of just coding your gadget to Facebook, you can have it run in 10, 100, maybe more places.

2) Rich Profile pages. Instead of having to reverse engineer Facebook, lots of site that couldn't afford to do that can now have profile pages as rich as Facebook's

3) Application and Mashup ecosystem. Like the Twitter ecosystem but times 10. Having the same APIs available at lots of sites opens up new possibilities for combining and adding value to the data.

There's a 4th side effect which is to make OpenID, oAuth and possibly AuthSub very common. It should give a big kick to adoption.

And now for the speculation.

If you want to get really radical, think of a site like Flickr or Twitter. I can just about imagine Flickr written with no account support at all and as an embeddable OpenSocial gadget. That feels like it has lots of implications. Almost all the gadgets so far have been small utilities. But try and imagine a rich application with multiple pages running inside the OpenSocial framework. It wouldn't need any login or account or profile management. It can get that parasitically from it's host container. It might well have a secondary website that aggregated all the activity from all the containers. This has many implications. Here's a couple:-
- How the hell do you monetise it?
- This might well allow Containers to promote best of breed in certain application areas instead of building a poor copy themselves. Think Discussion groups or Event management.


[ << Ecademy now has support for the OpenSocial People Data API ] [ Communicate >> ]
[ 10-Nov-07 8:09am ] [ , ]