I see this trend towards cloud computing as very much a web 2.0 phenom (maybe more like a 2.5, given how recently it has built up to the point of buzzworthiness). After all, the cloud wouldn't be necessary if all of these remote silos hadn't been created in the first place. The cloud is a band-aid.

Yes, it's very convenient for the developer to code for one computer as opposed to tweaking it for millions of exceptional ones, but I think that points out the larger problem of end-computers. It's also important to note that it's easier for the developer, but ultimately their ego-centric needs [and note I use this in the Jungian context, not the connotative insult: ego-centric here refers to "my POV prevails and is correct by default"] are superfluous in a consumer-oriented market. The customer is king, not the developer.

I see ajax as yet another shoehorn of a webpage into an app, and that it's bassackwards as well. Gears is wonderful in theory (MySpace is already using it for their email system – strangely progressive of them). Everything should be “geared” so that I always have a local (primary) copy of my data, and that it's only shared when I need a remote and objective “space” to do something that directly querying my peers cannot achieve. There are very few things that can't be done p2p.

In a p2p architecture, there's no need to centralize the database (there's only one database – mine, and temporary ad-hoc ones on the server that don't need to be maintained), code updates are distributed effectively, and you don't need to worry about overloaded web servers when most data is pushed and pulled amongst the users directly.

There's nothing about a p2p architecture that prevents a user from being able to access data from anywhere, if you use a remote storage service such as xdrive or the mythological gdrive that acts as your p2p proxy.

Need remote app processing because your iPhone can't handle the load? That's where HaaS comes in, and I think Amazon is showing its viability.

This model satisfies one basic theoretical concept in both database and hypertext design – write once then refer (backups notwithstanding). My problem with such groups as DataPortability is in the name – data isn't supposed to be portable! Standardized, yes. Cached, sure. Copies made? Nuh-uh.

The real key to future development of the web is recognizing when something is more appropriately done in a subjective or objective manner. That is value added, not value hyped.

If I give a site access to my data, I've lost control of it – period. One just needs to look at the constant violations of privacy with Facebook as the prime example.

There's a fundamental disconnect with all things Web 2.0. A reversal of when the subjective and the objective (in the Jungian context) is appropriate. There are only a few reasons why one needs the “objective” (shared/compromised/external – websites) when the Internet is based on subjective (private/personal/internal) networking. There are few things that web 2.0 companies do today that couldn't be achieved by distributing the code and using P2P technologies amongst the users.

Of course, they wouldn't be able to monetize that. Hence why it hasn't happened.