We held a session at RootsCamp DC on Google App Engine in which I was a total fish out of water, and I've been meaning to drop some notes here on why. I walked into the session apprehensive about the seeming trend towards Web 3.0-as-platform, but the programmers, vendors, and advocacy staffers assembled in this meeting were (justifiably) excited about the free, easy, and scalable platform that Google had just offered up. There's no denying that Mountain View is now offering developers something very tempting.

But at the risk of sound alarmist, I'll say that Google App Engine should be approached with a hint of caution, and here's one concrete reason why. If what I learned in that session and from reading up on it afterwards is accurate, one of the trade-offs in building applications using GAE is that you have to write the code that Google wants you to write. It makes sense that Google should have developed a style of programming that both protects their enormous investment and makes use of the distributed nature of their server farms. But it does mean that generations of coders are going to learn how to code for the Google cloud. Creating a whole mess of ready-made employees is a boon to Google, of course. But even if we get less Orwellian with it, we have to see that as Google Apps Engine popularity grows so does Mountain View's already considerable power to shape the industry.

When we did start discussing some of the potential negatives of GAE in that RootsCamp session, one of the participants said something like, "yeah, but Google just makes everything so easy." Well, there was also a reason why that apple in the garden of Eden was so juicy and sweet. ;)