DISQUS

I also believe this is a step in the wrong direction. Most serious programmers want to use high level fourth generation compiled programming languages. Java is probably the best one for highly available serious web applications. However, the Facebook Java client library was bad at best. In the applications that we wrote, we wrote our own library. It isn't that difficult and it allows you to have control of the entire application. I guess their thought is if someone is capable of programming scalable Java applications, they can write their own libraries. So it doesn't hurt that they dropped support of the client library, but it is a bad sign for the ecosystem. It is the serious programmers that are going to drive Facebook applications to the next level and Facebook seems uninterested in building the ecosystem necessary to support these companies and developers.

With all new major technology introductions, it is overhyped in the short term and underhyped in the long term. This is exactly where Facebook is heading. I think the impact long term is still going to be profound, but we are all just getting used to concept now and the ideas don't seem as buzzworthy. However, they are still very radical concepts to the general public.

I think you are absolutely right on. This is an age old problem between a platform and application. The general rule is that the platform will generally swallow up applications that are "Too Close" to the platform. The key for an entreprenuer is to inovate a true application rather than a small feature on the platform. Also, I predict that Facebook will eventually write "Killer Apps" for Facebook. This is much like the "Killer Apps" for Windows(Word, Excel & PPT) are written by Microsoft. As an app developer, the key is to understand this and to truely differentiate your offering. It is time for the Facebook application developers to develope real applications rather than trinkets that can easily be replaced.