CBDI Forum CBDI Web Services Roadmap. Guiding the transition to Web Services and SOA
 
eLearning

A Web Services Maturity Model cont'd...

Introduction

Every parent is familiar with the sound of the tiny voice piping up from the rear seats of the automobile "when are we going to be there?" Many are starting to ask the same question of Web Services, which are on the face of it, taking an inordinate amount of time to come to anything like widespread acceptance. Believe it or not it's nearly three years since Web Services first came on the scene, and four since XML-RPC activity signaled the paradigm shift. Two things are clear - first Web Services are far from mature by any measure, and second we have a long way to run before we reach anything like maturity.

You might well ask, "What defines a mature state?" Also "What happens after we reach maturity?" Although, as our recent survey confirmed, many organizations are adopting Web Services, equally many organizations are deliberately adopting conservative, mainstream adoption policies, and currently staying out of the Web Services market. In a difficult economic climate, why take a risk if you don't have to?

Yet Web Services are a little different to many other new technology trends. Web Services are clearly evolutionary insofar as they enable wrapping of existing application functionality, and provide early ROI based on improved reuse and better structured applications. And with the standards process very obviously following a course of layering complexity upon complexity, it is clear that a prediction of concept maturity based on the standards would be 2005 at the earliest.

Is there a risk that Web Services will fail, and be superseded by something else or simply swept aside by some new technology trend? Frankly this is about as likely as a collision with a stray interplanetary body. With the hegemony of Microsoft and IBM driving Web Services at the core of their business strategy, this is a technology that is going to run and run. So the questions to ask are when do you jump on the train and why?

And here we need to consider the relative maturity of both technology and business.

Technology Maturity

If we look at Web Services as a set of technologies, then we should expect to see an Atlantic breaker pattern, where rolling waves are continuously breaking on the shore, with something resembling predictability and regularity. Because the evolution of Web Services is fundamentally driven by the standards process, we can forecast when the waves of security, reliability, management etc will happen. Although there are competitive and commercial games being played with the standards process, it is to everyone's advantage that there is universal buy-in to the WS protocols, and therefore we can be less concerned about tactical battles, because the outcome is pretty much a given.

With most new technologies there is an observable model where early use of any new idea is perceived as an extension to existing practices. So the first automobiles were designed to look like carriages without the horse. Mobile phones provide(d) text messaging facilities that use the numeric keyboard irrespective of the difficulty of using the UI.

Although the use of Web Services for better integration is now becoming widespread, the reality is they are simply another layer on top of the infrastructure that already exists. This is particularly true in the Java environment, where there are so many layer mappings - from UML, to relational, from XML to objects and back again, and XML to code, that it's not surprising that organizations are going slow, unless there is a real imperative. In an upcoming report we look in detail at new XML based implementation technologies such as X# and Water, which clearly have the potential to make large swathes of the current development complexity redundant. So while the technology world waits for the WS standards to catch up with expectations, we might just see a major advance in implementation technologies happening concurrently, which will deliver massive simplicity into the services "delivery" environment.

Business Maturity

In Richard Veryard's recent CBDI reports on Modeling for SOA (CBDI Best Practice Reports Feb 03 - Modeling for SOA, Apr 2003 - Worked Example) he has described the opportunities that are going to surface as we achieve the reality of trusted, and ubiquitous service interoperability. In particular he has emphasized the importance of looking at mundane business processes and services in a new context, for example looking at the service ecosystem as the scope for business design. What we used to call reengineering before it became over hyped and unfashionable!

An equally interesting perspective to examine is the way that business may be transformed by service thinking. One of the more interesting reports that I have personally researched over the past six months was about BT's authentication service (CBDI Product Report Apr 2003 - The BT Authentication Service). In this CBDI report I discussed how BT is working on a long term program which aims to deliver a pervasively used authentication service. The really key point here is that the technical delivery of the service is literally trivial in comparison to the business task. Although BT is partnering with a company highly experienced in authentication and personal data management, they have a huge task to persuade initially businesses and subsequently individuals to change their customary practices in relation to personal identification. This is nothing short of reengineering on a grand scale. What's important to note is that BT is already embarked on this program, long before the technologies are anywhere near mature, and is focusing on business led product design, with the clear intent to converge with the maturing technology.

Web Services Maturity Model

So how do we rationalize where we are in the overall march towards a service oriented world? In our work we advise the importance of pacing activity in line with product maturity, whilst at the same time developing and working towards a longer term vision. The BT authentication program is clearly a good case study.

In Figure 1 we offer a very simple model to assist communication and planning. This model is not intended to be definitive or precise, rather a rough aid to understanding what's going on. For example if the columns don't all match up for all the rows at any point in time don't worry unduly, it's the principle of phased progression that's important, and the need to prepare and manage through it.

Figure 1 - The Web Services Maturity Model

Previous Page Next Page
CBDI Knowledgebase

  © Everware-CBDI Inc 1999-2008