There’s one topic that never seems to die out when we’re talking about Web 2.0, and that’s the definition of Web 2.0 itself. Simply put, people don’t know what it is.

The latest trend from the experts is to either dismiss (some aspects of) the term entirely, or to exclaim that Web 2.0 can be equated with innovation as such, or the web in general.

Another recent attempt at a Web 2.0 definition brings up a good point. The author does try to distil the essence of Web 2.0 by saying that “Web 2.0 is about doing stuff on the Web that can’t done in any other medium“. The idea sounds attractive, but as a definition, it falls flat. It boils down to “A is everything that non-B is not.” Many questions instantly arise: what are these “things” that we cannot do in any other medium? And why can’t we do them?

Finally, no discussion about Web 2.0 would be complete without mentioning Tim O’Reilly’s compact Web 2.0 definition. Tim covers all the important aspects of Web 2.0. It’s hard to substract or add from it. But his definition is not without problems. It relies on a number of terms that require a definition of their own, like “architecture of participation“. Also, it contains some phrases that are too vague to be a part of a definition, for example: “going beyond the page metaphor of Web 1.0“.

Tim himself says he’s not fond of having to define Web 2.0, and he has a good reason for it. The problem with Web 2.0 phenomenon is that it is a multi-pronged beast. If you take all the main aspects of Web 2.0, you’ll find an application that features one subset of those aspects, and you’ll find another one that features a completely different subset of aspects. So, the problem here is finding the lowest common denominator; the absolute minimum of features some service or application has to have in order to fall under the category of Web 2.0.

The old vs. the new web

But if web 2.0 is the same as web 1.0 - as many now claim - why didn’t these new patterns of usage emerge right away? Why are we noticing them only in the past couple of years? It is because it took some time for the number of internet users to reach a critical mass and for the power of networking to start working. The internet is not just a bunch of computers connected into a network. It is - to use an ancient phrase - more than just a sum of its parts. Not only does it enable users to communicate and exchange data; there are certain ways of manipulating data that are enabled by the Internet - for example, mashups. This is - I presume - what the author of the article on ClickZ meant: the Internet creates new possibilities for data manipulation, and whenever you use them, you’re basically using Web 2.0.

So, how do we define where the ‘old’ web stops and the new web - Web 2.0 - begins? The answer is - we don’t. For example, web pages, underlined by the HTTP protocol, aren’t just a bunch of static text and images anymore - they’ve evolved and can contain what are now called RIAs (Rich Internet Applications). So, a clear-cut barrier between Web 1.0 (if I may call it that) and Web 2.0 doesn’t exist. This solves one of the many pitfalls of Web 2.0: it is not a new phenomenon, completely separate from some other part of the internet. It is simply the utilization of possibilities that only Internet can provide.

Finally, Web 2.0 defined?

So, why do we intuitively know what Web 2.0 applications and services are, but are never quite sure what separates them from the old ways of using the web? It’s because web services and application don’t always use all the possibilities that the Internet provides.

You can use the web in this way, too, but you’re not using all that it can do.

We can roughly divide the entities on the web in four categories: the users, the data they produce, their computers, connected to the internet, and the connections between the three. In general, web applications which we would probably refer to as web 1.0 are those which disregard the last part of the equation: the connections. But when we define Web 2.0, we’re not defining all that it necessarily is - we’re defining all that it can be. Thus, the definition is as follows:

Web 2.0 is a data manipulation system, consisting of individual machines connected to the internet (or any large-scale network), their users, data these users produce, and the connections between those three.

When you start harnessing the real power of networking - the connections between all the entities on the web - then you’re using Web 2.0

The important thing to notice here is that Web 2.0 is largely not possible if the applications don’t use the web as the platform. If the users create data in an online application, and this application stores this data online, and in such a way that it can be accessed and used by other users and/or applications, then we’re talking about real networking.

Since the web is not really a machine, it cannot be called a computer, but that’s what it essentially is. When you hear the word ‘computer’, you should not be thinking about a keyboard, a case, a monitor and a mouse. A computer is a machine for manipulating data according to a list of instructions (source: Wikipedia). It is that simple. When you open Flickr and start browsing images, you’re accessing the data of many different users, that was uploaded using many individual computers. Connections between the users and between the data sets are as important as the data itself. This is the part that no desktop platform can provide.

In that sense, Web 2.0 is a new kind of computer, while Web 2.0 applications are simply applications running on it.

These applications can use all that Web 2.0 has to offer: sharing, networking, mashing up, remixing and continuous updating, or they can only use some of it. That’s what “using the web as a platform” means, and that’s the key to Web 2.0. However, the web as a platform is young, undefined, volatile and not nearly robust enough at this point. That’s why trying to create an all-inclusive definition like O’Reilly’s is an impossible task, as new aspects of Web 2.0 will keep appearing in the future, when new protocols and technologies, which will make the web more platform-like, emerge. By the way, this is also why I think that any attempt at defining what Web 3.0 might be like is still pointless, as Web 2.0 has a long development ahead.

What I’ve tried to show here is that the Web has evolved, and that this evolution is recognizable in a clearly definable set of phenomena which we often refer to as Web 2.0. Is this the best way to call it? Probably not, but it is, in a sense, accurate: it is the logical progression from simply having a bunch of computers connected in a network to actually harnessing all that this network can offer.