YOUR FEEDBACK
andy.mulholland wrote: intriguing !!! We have full scale 'Mashup Factories' in Chicago USA and Utrec...
SOA World Conference
Virtualization Conference
$300 Savings Expire October 17, 2008... – Register Today!

SYS-CON.TV
TODAY'S TOP SOA & WEBSERVICES LINKS


Building SOAs That Benefit Business Users
Matching user needs to the IT benefits of moving to Service-Oriented Architecture

If Service Oriented Architecture (SOA) is all about business agility, then why does the focus continue to be on how services will be built, deployed, and managed by IT instead of how they'll be consumed by business users?

How will SOA services to be rolled out in your organization? Through a Web browser or rich Internet application? Will these Web applications totally replace the applications your business users are currently relying on? Or will they simply add to the growing number of applications that your business users need to access on a regular basis?

Enterprise architects have multiple technology options to choose from as they implement an SOA. However, designing it with the end user in mind can help dictate choices that will speed up application integration, increase overall flexibility, and result in a more productive workforce. By creating a New Enterprise Desktop, architects can transcend the latest SOA messaging, governance or security fads and focus instead on what really drives the successful enterprise - a superior SOA user interface.

Today's Desktop Environment
Let's start by examining the current environment for most of today's business users - the desktop. The reality of the current desktop environment generally consists of the following challenges:
1.  Users must interact with multiple applications. A good example of this is the call center industry where, through an informal survey of our customers, the average agent needs to interact with nine-10 applications on a daily basis to resolve customer issues.
2.  The more applications used by business users, the more an enterprise's bottom line is affected by reduced productivity and higher training costs.
3.  Most applications today are not integrated. Despite all of the investment in integration technologies, the most common form of integration today is still manual copy-and-paste.
4.  Most business processes require humans to complete tasks manually.
5.  Application functionality is generally fixed and limited to what the original developers created even if it doesn't meet user requirements now in terms of functionality or ease of use. Major change requests queue in your IT backlog or in vendor product roadmap feature request lists.
6.  Users are increasingly interacting with applications controlled outside the organization. SAAS applications are good examples of this. More sophisticated partnerships and outsourcing arrangements are also becoming increasingly common.

The promise of user productivity from newer applications generally isn't realized because new applications rarely replace existing applications or streamline processes for users. They simply add to desktop complexity, which is driving the need for the New Enterprise Desktop designed around end-user productivity.

Despite the best intentions, today's business users are required to interact with many different disparate applications to perform daily tasks

SOA in Today's User Environment
The promise of SOA is that applications will be broken down into a set of reusable services creating a more agile computing environment; one in which new services can be created and deployed more quickly and easily. But even the most optimistic of us understand that full-fledged SOA is a long journey, not a near-term reality. In addition to automating services, SOA architects must define processes around those services that secure, govern, and manage their SOAs. In fact, as "service reuse" becomes a mantra, middleware vendors are rolling out vertical industry roadmaps with standardized components that run on top of an SOA platform.

The near-term reality is that services will gradually be rolled out but not as a complete replacement for existing applications. Services will be incremental to existing applications in most environments.

From a user perspective this means that new services will likely be made available via a browser interface adding yet another "application" to learn and use and likely fall into the copy-and-paste category along with other applications.

At our most basic level, business users really want technology to help us do our jobs more quickly and easily. Simplicity and productivity are paramount. We want as few user interfaces to learn and use as possible and yet we need new richer and more integrated functionality including that derived from SOA.

The Frontlines of SOA
How will SOA-based services ultimately be consumed by users in your organization? The most common thinking is via Web applications or "enterprise mashups" but that's only one option and likely not the best solution for most enterprises. The three most likely options are:

1.  Services delivered through Web applications or mashups. In this scenario, the browser becomes the new operating environment. But how long will it be before these Web applications or mashups totally replace your existing applications throughout the enterprise? Is that even your plan? The most likely scenario is that Web applications will increasingly deliver business functionality but need to interact and integrate with other enterprise applications, many of which will still reside on the desktop. Web applications will remain one of the application platforms that deliver functionality to users; not "the" application platform. The advantage of such an approach is that there will be plenty of choice in terms of Web application frameworks in which to deploy your services. The downside is that the Web application is only one of the many UIs that a business user must interact with to do his jobs. You could argue that such an approach actually reduces user productivity.

2.  Extending existing applications with new Web Services functionality. In this scenario, new services are added to existing applications, so users benefit from new functionality while still interacting with their existing UIs. The main advantage of this approach is that it enables IT organizations to roll out new services with the least impact on business user productivity. Such an approach requires a technology platform, such as the OpenSpan Platform, that can be injected into the memory of running applications to execute the desired automations, such as calling a particular Web Service for a given keystroke or mouse click.

3.  Building new thin-client composite applications that reside on user desktops. In this scenario the composite application isn't limited to just Web applications but virtually any legacy or desktop application. Such composite applications mesh well with SOA architectures in that they view supporting applications as nothing more than a set of services, bits of functionality. Composite applications let developers identify only the relevant functionality in a set of applications and deliver that functionality to users in an easy-to-consume format. The Afni example in the next section illustrates both the architecture and benefits to this approach. A major advantage is that composite applications can dramatically increase user productivity and reduce training requirements for new employees because, unlike mashups, they tend to reduce the total number of user interfaces with which a business user must interact. They also provide a mechanism for rolling out new services more quickly. IT organizations can then replace or upgrade business applications at any time without necessarily impacting the user experience on the desktop. They can also gradually roll out new SOA services, replacing older legacy integration points again without negatively impacting the user.

Now, it's likely that your enterprise is already heading down the path of option one. The developer side of you is probably thinking that SOA = services; integration = middleware and ESBs; GUI = Web applications. Are you still searching for a framework for building Web applications? Building a user interface shouldn't be a framework since that would imply that you're going to put serious business logic in the wrong place. User interfaces should be built with user interfaces and not a lot of code.

The front-end of SOA should either be your existing UIs or, more likely long-term, a composite application. A composite application will let you pull together documents from Google, customer data from Salesforce.com, partner data from partner Web Services, as well as other enterprise data from your existing legacy applications. The future of SOA requires a composite interface, a New Enterprise Desktop, and not additional composite middleware applications.


About Joe McGonnell
Joe McGonnell leads worldwide marketing activities for OpenSpan. Prior to joining OpenSpan, he spent three years leading marketing for JBoss, a highly successful open source software company eventually acquired by Red Hat in June 2006. Joe was the first marketing hire at JBoss and helped to build a highly successful model and marketing organization that drove awareness for JBoss products and services and helped to position JBoss as a leader in the open source software movement. Prior to joining JBoss, Joe spent more than ten years in various product management and product marketing roles for Hewlett Packard, Bluestone Software, and Checkpoint Systems.

SUBSCRIBE TO THE WORLD'S MOST POWERFUL NEWSLETTERS
SUBSCRIBE TO OUR RSS FEEDS & GET YOUR SYS-CON NEWS LIVE!
Click to Add our RSS Feeds to the Service of Your Choice:
Google Reader or Homepage Add to My Yahoo! Subscribe with Bloglines Subscribe in NewsGator Online
myFeedster Add to My AOL Subscribe in Rojo Add 'Hugg' to Newsburst from CNET News.com Kinja Digest View Additional SYS-CON Feeds
Publish Your Article! Please send it to editorial(at)sys-con.com!

Advertise on this site! Contact advertising(at)sys-con.com! 201 802-3021


SYS-CON FEATURED WHITEPAPERS


ADS BY GOOGLE