The web as we knew it in 1995 has already largely died. Out of the ashes has arisen a second incarnation and we are currently on the verge of a new reality, Web 2.0. While there is no one definition, Web 2.0 is perhaps best described as the migration to the web as a platform spanning all connected devices, coupled with a specific set of patterns. Web 2.0 has many components, but it is generally associated with a class of web applications that harness the intelligence, data, and actions of their users to create value (iconic Web 2.0 applications include Flickr, YouTube, and Amazon). While many are looking to Web 2.0 to solve the problems of yesteryear, the mass migration is creating a new set of problems that must be addressed.

This article is divided into three parts: an analysis of the web today, an analysis of what has already died or is dying, and a look forward at aspects of Web 2.0 that are creating problems and will likely die in the next five years.

The Web today: Introspective analysis

The new model and core set of design and architectural patterns of Web 2.0 have enabled companies and individuals to migrate to the web as a platform. This has stimulated new business models, new interaction patterns, and most importantly, opportunities for disruptive innovation.

Web 2.0 cannot be defined as a static architecture nor should anyone attempt to do so.

Some of the knowledge of Web 2.0 can be captured and preserved as a set of core design and architectural patterns that are shared by a group of new web sites, portals, and web based applications widely considered to embody Web 2.0. (The O'Reilly publication "Web 2.0 Design Patterns", ISBN 0596514433.) Underlying themes of participation, collaboration, community, co-creation, connection, trust, and openness are the basis for many of these patterns.

Web 2.0 ideas can help create value for all by driving innovation within organizations, improving the way teams work, and tapping into the ideas and talents of all employees and constituents. Developing expertise in the Web 2.0 space creates opportunities for governments, businesses, and ordinary users to create, reuse, and distribute information.

Partners and customers want to use Web 2.0 to create value through communities, collaborations, co-creation, and connections; businesses want to use it to improve the way they work, innovate, engage, and deliver to clients, partners, and employees.

To better understand the death of the web as we knew it, and also look ahead at the problems facing today's Web 2.0, it helps to take a step back and review the evolution of the Web.

The old Web

The first generation of the Internet was largely based on the client-server model, a two tier model in which clients initiated interactions with the servers to manage or get resources. Many organizations simply extended certain pieces of information to servers so they could be accessed by a client via the httpRequest() method. The requestresponse interaction patterns were over after the initial request was completed and users who received the information had to often complete the remainder of their processes via other means. An example of this might be a user who gets a travel brochure from an online agency, yet has to pick up a telephone to complete the holiday booking process.

Figure 1. The Client-Server model of the first Internet.

A new model has emerged for the Web which has facilitated many of the Web 2.0 patterns. Shown in Figure 2, this model reflects Web 2.0 as a platform that connects capabilities and users for the entire duration of a process -- from start to completion. It is reflected in implementations in several different and disparate environments. In one implementation, an enterprise shares corporate information with its customers. In another implementation, the model is applied to a peer-to-peer (P2P) file-sharing application. It is likewise applicable beyond the Internet itself. Software developers could use this model to expose the functionality encapsulated in a single class to other software components via a defined interface throughout the specific environment (connectivity/reachability) in which the interface exists.

Figure 2. An abstract model for Web 2.0 architects

Web 2.0 is a platform used to span connected devices via services, yet it is also much more. As more and more types of devices are connected to the Internet, the services have to be more carefully thought out in terms of their architecture, implementation, and descriptions. The client tier, expressed in client-server models, has been split in this model to emphasize the two distinct aspects of the client in the model:

• the applications and runtimes, and
• the users that are the ultimate targets of most interaction patterns.

Because Web 2.0 is largely about humans as co-collaborators in the success of a process, the model must reflect their importance and inclusion.

What has already died

The old days of client-server technology have not yet died completely, however no new business models are likely to be based on such a simple model. Today's Web recognizes that users are less loyal and less patient than they were in the 1994-2000 era. When designers and deployers of web applications allow only part of an engagement to occur on the Web, they risk losing the majority of the users.

So what specifically has died in the last five years?

1. Simple client-server, request-response architectures. These simplistic interactions have been partly retired in favor of applications that know something about the users. Most new interactions use the model above, in which users (often humans but also other applications) interact with some capability via the Web.
2. Not knowing much about your users. The first version of the Internet was based on idempotent, largely anonymous user interactions as noted above. The new reality is that web application users demand a more engaging and contextual experience. Users are seeking more specialized content. Examples of the new model include Google's customizable homepage and Facebook.

3. Non-rich user interfaces. Boring graphical and utilitarian user interfaces are dead. Many companies have expressed that the single largest challenge they face is finding a way to graphically present information to users in a manner they can comprehend. Richer interactive applications are the key to this demand. These new applications have moved beyond straight HTML to include Ajax, Flex, Silverlight, Java FX and Flash.

4. Non-open applications. Web applications without APIs and other interaction mechanisms that people can use are also on the scrap heap for 2008. Users want more control over their interactions. For example, Twitter, mostly known for its SMS capabilities, is better described as a device-agnostic messaging system.

   Users can choose from a range of consumption and sharing methods, including the Web, instant messaging, custom clients, alternate web sites, and SMS.

5. Platforms not built in alignment with SOA. Infrastructures that depend on tight binding and non-repurposable capabilities risk becoming the next victim of change. Service Oriented Architecture, or SOA (as defined within the OASIS Reference Model for SOA), is an important facet of most new web applications including those that use patterns such as Software as a Service (SaaS) and Mash-ups among others. Service Oriented Architecture, in this context, is an architectural paradigm for matching needs and capabilities amongst disparate domains of ownership. SaaS is software distributed as a functionality, as opposed to software that has to be installed on every machine.
6. Non-collaborative interactions. Users want the ability to collaborate and contribute. Applications that are broadcast-only and lack at least minimal user interactions (for example via comments, tagging, or mashing up content) are destined for the scrap heap too.

The time for applications burdened with these problems has already passed. Now let's look at some things going on today that will die in the next five years.

What will die tomorrow

While Web 2.0 as a movement has improved the collective experience of the Internet, there are also some red flags that developers and users alike are raising today. For example, how many times have you set up relationships with the same people in Plaxo, LinkedIn, MySpace, Facebook, MyTopFriends, Spock, Friendster or another social networking site?

Let's examine a few things that are likely not going to last the next five years (or even three).

1. Duplication. As users, we are asked to enter the same information over and over again. Whether it is identification, personal details, or declarations of relationships, we should not have to re-enter it over and over again. Why should we have to re-declare the same business contacts in both LinkedIn.com and Plaxo.com? Wouldn't it make more sense if we could simply export them from one and then import them into the other? Or better yet, let us declare them once in a large scale open framework and have both Plaxo and LinkedIn pull that information out.

   Many companies resist such a solution because they fear losing users to their competitors. There is currently a movement underway to promote data portability that seeks to enable users to share the data with whomever they choose.

   As policy discussions continue there has been progress on the technology side.

   OpenID has the potential to become a universal identity and OAuth is an emerging standard that lets sites share data safely and under user control.

   In time someone will take the lead and first build an ontology (a shared conceptualization) of the human-to-human relationship types, then use it to architect an open framework that individuals can use to manage their relationships once and federate that content to other social systems where applicable – all while maintaining control of their own information.

2. Lack of openness. Walled gardens are starting to crumble today as more and more people migrate from web sites and portals to platforms. When one studies the architecture of Facebook, it is easy to see that it is a platform rather than a walled garden. The leaders of Facebook have learned from the mistakes of previous attempts and have allowed developers to extend the core platform. What is the win? Those same developers go out and promote the platform as they try to monetize their extensions.

3. Simplistic approaches to semantics. The idea of the semantic web is a grossly underappreciated issue and the kindergarten efforts to build it using tags are not going to scale. The semantic web is a nirvana in which rich metadata allows both human and machine agents to locate resource much more accurately. Semantics without the guidance and framework of an upper logic (ontology) are doomed in many ways. The current approach assumes that if we tag something, others can understand the semantics. Some have even labeled this "Web 3.0", which only demonstrates they have not grasped the fundamental concepts of Web 2.0.

   What is required? A framework whereby the first order of logic can be used as the guiding principle for the semantics domain. The effort has to scale to be a multilingual, shared conceptualization of the Web and the information on it that is grounded in references to reality. Think about this for a second. Can anyone really understand what "spicy food" is without actually tasting it?

4. The forking of the Web. Standards still matter. Web designers do not want to have to develop six AJAX applications for two different versions of three browsers because there is no agreement on how to implement a standard (or standards). This happens when the people who implement standards no longer care to design and implement them in ways that achieve true interoperability; it results in forking the web.

How can we fix these problems?

For now it is more important to understand who is going to fix these problems than how they are going to be fixed. Each of us is part of a global community. Where we see things that are broken, we each have a duty to the community to help with the cure. The web as we know is standing in the ashes of the old web. The new realm we are entering has problems that are going to require us, as a community, to agree on a way forward.

Over time those who do not adapt to the new ways of thinking (usually evolved after many experiments are tried) will slowly fade away.

This work is licensed under a Creative Commons Attribution-Noncommercial 3.0 Unported License.

About the authors