If one were to look at the world of networking as a layercake of virtual machines, you can probably see that the world of web applications has become surprisingly complicated. There is so much to know and so many interlocking specifications to pay attention to that we shouldn't be too surprised if web users slowly begin to feel overwhelmed.

I am not sure why in the age of usability and information architecture we seem to have more interaction mechanisms than your average mother-in-law (please don't spam me for this - and I am apologizing to all mother-in-laws forthwith). We have RSS, Atom and a few of their relatives. We have SOAP, REST and various sub-protocols. We have a jungle of authorization and authentication mechanisms. The faults of http authentication were known from the point of their conception and are now being augmented in practice.

Still, I feel we are searching for a good idea that does not exhaust itself in commercial applications. Luckily, we are not as naive anymore when we thought that CORBA would be the be-all and end-all of commercial distributed computing running on something resembling TCP/IP (or rather, trying to reinvent it). So, we know that these protocols are fun and games, but they are also insecure, and simplify matters for a programmer, not for a user.

I look at flickr and facebook and my heart sinks, not because flickr is a bad thing - it is extremely convenient - but somehow I am not as overwhelmed by my ability to store pics on flickr and then being able to use some Google API to give someone an idea where this pic was taken. Great. We used to call it geography and we have done it by hand since the invention of photography since the 1840s.

Web 2.0 automates some data correlations. That is quite useful, but it isn't any great news to anyone who has been around, say, GIS systems, or anyone who has handled 3D data. Sure, it is rather convenient to have it available using web interfaces of some kind, or even have it represented on Second Life. But I still don't see the revolutionary aspect of it, except that many more people can play with it than, say, before the web became commonplace in the late 1990s.

In all, I would be prepared to call web 2.0 cultural evolution: we have democratized data access further and we have given visual data structures a life on the web. That is not a small achievement. The same applies to audio data and other "multimedia" formats. I still believe we are fairly close to monks copying and illuminating theological and medical texts not knowing that the shack down the dirt road is spitting out hundreds of Bibles in weeks, using something called the printing press.

We still have a long way to go.