Agreed. The simplicity and elegance of the API is really great. A lot of sites have status updates + commenting + liking + feeds... and this essentially gives y

Agreed. The simplicity and elegance of the API is really great. A lot of sites have status updates + commenting + liking + feeds... and this essentially gives you a lot of that functionality for 'free'. You don't need to worry about the hard problems in regards to friend visibility, scaling the push model, etc etc. All the difficult data processing is handled on Facebook's end while a site becomes more of a lens/context/perspective on the data. Of course the cost of this is that Facebook gets your data. It's a bit of a Faustian contract (hope I'm using that reference properly). They'll save you engineering cycles and promote your content to your user's social graphs, you just have to give them your data. Sites really now have 3 options: 1) Ignore FB to their own risk 2) Embrace FB and give up control of their data 3) Use FB as a data mirroring and distribution mechanism - but still give them all your data. Choice 1 is a difficult decision to make... if FB becomes THE standard and you're not providing your users the experience they expect, maybe they'll go somewhere else. Of course this could be like Beacon where users hate it. Or it could be like Facebook Connect where it's cool and some sites leverage it really well, but hasn't change the landscape of the web THAT much. If you choose Option 3 you're still giving your data to Facebook. So now you're left with, is it worth the engineering cycles to ensure you don't get locked out of your own data. I think most people on here will say yes, but a surprising number might not. Now after thinking about it for a while, this will cause an uptick in Facebook Connect usage and integration, but I don't think it's going to fundamentally change the way sites are designed. Maybe in a year or two if Facebook continues its growth current trajectory.