I understand the intuitive appeal of applying versioning to the web, it gives a name to a phenomenon that is otherwise pretty difficult to summate. How else can you articulate the combination of web-centric interaction, web-only business, AJAX behaviour and socially networked and distributed application systems succinctly? Web 2.0 - or “Web as Platform” is a fairly universal way of tokenizing the phenomenon we’re all basically aware of, even if we don’t understand it.

The issue with that kind of nomenclature is that it fails to recognize a number of aspects which are fundamentally important to how the phenomenon is taking shape. While less important in the 1.0 / 2.0 debate it’s especially true for the 3.0 discussions which are (already) beginning to crop up.

Web 2.0, as Tim O’Reilly defined it in his (seminal) paper on “What is Web 2.0,“  was the web as a central platform instead of extension. The paper  is certainly a landmark in the general consensus of what Web 2.0 promises and is just starting to deliver so I’ll use its definition, which is not without controversy when discussing Web 2.0. The fact that I need to disclaim this simply goes to show how confused everything remains.

Web 2.0 is esthetically represented via AJAX, which is the technical manifestation of the phenomenon. This is what many believe IS Web 2.0, Eric Schmidt being a big proponent of this techno-centric history. Most techno-centrists believe that although Web 2.0 manifests itself in a much broader sense than simply the standard of markup used and a particular javascript method (xmlhttprequest()), it was these enabling tools that allowed all the other facets to flourish - without it Web 2.0 could be at best described as web 1.5.

This is a seriously deficient explanation for a number reasons. The most obvious are that it’s anachronistic and implies a dependency that doesn’t exist. AJAX (asynchronous (use of) javascript and XML/XHTML) is not necessary to make a Web 2.0 app - just as elevators are a signifier of modern architecture a building without an elevator isn’t necessarily NOT modern. In fact Web 2.0-ness is a much fuzzier quality than use of AJAX, which rankles people with engineering or comp sci degrees whose jobs are absolutely dedicated to precision.

Web 2.0 is the view of the web not merely a means to some external end, but as an entire economy and society unto itself. Services, products and information can be commodities that are traded and purchased without any value chain leaving the internet. This is what separates companies like 37 Signals from Amazon and Microsoft, web 2.0 from web 1.0. William Gibson conceived this as cyberspace, an alternative reality that we perceive as equally viable to our lives as the real one.

Equating AJAX with Web 2.0 is an egregious example of post hoc ergo propter hoc (my Classicist sister would be proud!). Due to the sequence of events it’s intuitively appealing to infer causality - but there’s a huge difference between causality and facilitation.

AJAX lowered the barrier to entry with regards to interface. The web suddenly became just like desktop apps, sometimes even more usable. The tractability of what was feasible from a UI perspective exploded the ability to create new and significantly more meaningful interactions. Some believe that this fundamentally caused everything else - just as some theorists believe(d) the discovery of a large variety of pigmentation triggered the Renaissance painting. Unfortunately, they ignore the parallel activities that gave force to either Caravaggio or David Hansson (creator of Ruby on Rails).

Social networking, social distribution, and just as importantly the user created media explosion are huge part of the 2.0 impetus. Cheap digital cameras made sites like Flickr possible. $40 web cams, MiniDV and cheap broad band enabled YouTube. Frustration with pay-sites and paucity of good reference content enabled Jimmy Wales to bring Wikipedia from Nupedian obscurity to one of the most popular websites on earth.

Social networking was only made possible through the steady pervading of voyeurism, brought about through things like blogging and high-profile dot-com oddities giving people that lottery-ticket incentive of fame potential to get in.

The capability of being connected to the internet 24/7 - via wifi, cheap dial-up and mobile access gave us (for better or worse) things like Twitter.

Ruby as a language is older than Java - yet it is attributed with kicking off this Web 2.0 by way of Rails. The importance of Rails is not Prototype (the AJAX component of Ruby on Rails) it’s the fact that it rendered trivial Web 1.0 sites. Anybody with a decent book on RoR and a computer could recreate Amazon in an afternoon - so the standard for innovation was raised. Serious developers needed to step things up or get left behind.

The only rational conclusion that can be made with all these observations is that Web 2.0 wasn’t “triggered”, it was the simultaneous confluence of a great many other phenomena - some technical, many social.

Web 2.0 brought the web up to speed with our expectations of desktop computers - it finally caught up to 1994 with regards to interface, capabilities, portability etc.

There’s a big problem with versioning these things because it implies a sense of sequential progress that simply doesn’t exist - for that I think Tim O’Reilly needs to re-jig his thesis. For versioning systems in general there is an implied sequence to a version, not an implied chronology - which is really what we mean when we say 2.0.

3.0, according to yet another brilliant Timothy - this time Sir Berners-Lee, and many others is the semantic web. For web “historians” this is problematic - the semantic web is very nearly as old as the web itself. If HTTP and all it implies is the web, it could be arguable that semantic interchange of data predates the web as a technical goal.

The semantic web is an equally problematic definition for 3.0 because it’s difficult to square away with many other phenomena which are separate but otherwise bundled in. That is the “data cloud,” multi-device interchange, geographic awareness and ambient findability (as Peter Morville put it).

To describe in brief - the data cloud is the phenomenon of data no longer residing in a discrete resource. Your social networking information will be passed between the various applications that use it, rather than each network retaining its own data separate from the others. The Data Portability initiative is the highest profile organization pushing for this and its no coincidence it’s led by Tim Berners-Lee. In short in the data cloud, your data are everywhere and nowhere at once - much like the analogous electron probability cloud from which the data cloud gets its name.

Multi-device interchange is a pretty simple phenomenon with profound implications. No longer will the web be the domain of personal computers.  The smart phone will simply be “the phone,” normal “dumb” cellular phones have more or less gone the way of the dodo and the trend will push further and further toward computing mobility (rather than portability). Intel has been showing off its concepts for Mobile Internet Devices and these are just an obvious beginning. Amazon has shown us that even traditionally non-digital media are getting the boot with the wild-fire success of Kindle - an internet ready eBook. Laptop computers have overtaken the sales of desktops in North America and the size, range and power lifetime for devices is the primary drive of technological innovation (Moore’s law is less true now than it was 4 years ago).

Part and parcel with mobility is geographic awareness and geographic data. We’re seeing fledgling concepts for this in mobile devices with built in mapping technology and of course the still-flawed dashboard guidance consoles. Geo data will become more integrated into our experiences as we run up the learning curve. Digital cameras will “stamp” geo data onto photos (some do already). The Kindle will likely auto subscribe to local papers (just like switching to local radio stations). Trip and route planning will be more integrated into our mobile devices - including dynamic data like traffic, weather etc.

Another facet to this is ambient findability. This is arguably the coolest non-semantic aspect of the “3.0″ web. Just as Google is your resource for finding stuff on the web, and for most people is also the resource for finding locations (thanks to Google maps and GoogleEarth) it (or more agile competitors) will be your resource for finding “things” in the “real-world” or finding connections between real-world and web-sphere phenomena. This goes well beyond being a directory service like some really big yellow pages and becomes much more directed and personalized thanks to things like collective intelligence and the aforementioned multi-device interchange and geopraphic awareness.

Amazon helps you find a book on topic using a wide variety of data it has collected. Ambient findability will allow it to not only locate businesses or locations that fit in with what you’re looking for, it’ll connect your search and data with others and provide mechanisms for physical interaction. People on your friends list who will be at the same place will be alerted of your presence, car pools can be generated automatically without relying on the limitations of posted boards, the list of various services will go on and on as the web predominates as the method of information exchange - particularly asynchronous, fact-heavy information.

Personal possessions and products, thanks to RFID will become findable on the net. “Where did I put my sketchpad…? Oh well, I’ll just Google it.”

I might also note that the term is ambient findability not ambient searchability. Web 3.0 is probably the greatest threat to Google on the horizon because it deals pretty serious damage to search as the mechanism for discovery. I’ll save this topic for a different post, but suffice it to say either Google is cooking up non-search discovery technology (VERY likely) or is doing its best to keep 3.0 tech relegated to obscurity (also likely).

So, just as 2.0 isn’t a neat package of widgets neither is 3.0. In fact, 3.0 is much more problematic because it’s really two or three separate phenomena occurring at roughly the same time.

I for one think that the web is sufficiently large and complex to render sweeping generalizations or “trends” very difficult if not impossible to make without being entirely specious. It’s pretty difficult to make generalizations that apply to 60+% of the entire world without stretching things like taffy or simply uttering truisms.

Posted by Jeremy on Friday, April 18th, 2008 at 5:57 pm.