Under increasing financial pressure, organizations must identify opportunities to extend the life of existing infrastructure, pursue low-risk high-reward initiatives, and fully leverage existing technology. Extending mainframe applications to handheld devices achieves these goals given that:

• 70% of all business data is stored in mainframes.
• Numerous mission-critical sales force, field force, supply chain management, and asset management applications run on the mainframe.
• Reliable, stable, and secure access to the mainframe from handheld devices is available via XML.
• Leveraging the mainframe avoids costly e-business upgrades and reduces the risk of rewriting proven applications while generating significant ROI.

• A ready base of support staff - from operators to systems architects who are trained, experienced and available. No "learning curve" here
• Security capabilities in use by the most security-conscious organizations in the world
• Mature (thoroughly debugged) operating systems with all the centralized services needed to effectively and efficiently manage operations
• Hardware and software reliability that ensures virtually 100% availability
• Scalability and performance characteristics that meet the most demanding, high volume requirements

The vast majority of the data on mainframes today is supporting mission-critical applications. These applications run the gamut: sales order entry, bill of materials, distribution, purchasing, customer information, financials, payroll, etc. Exposing this data directly to those who need it anytime, anywhere, while maintaining all the strengths of residing on the mainframe, gives the organization that can do so a cost and competitive advantage.

Leveraging these applications, in place, reduces risk over purchasing, deploying and porting these applications to new technology. By leveraging the investment in existing mainframe applications and support infrastructure, the return on investment for new and enhancement-focused initiatives is maximized. The solution architecture described herein provides such leverage.

Extending Mainframe Applications through XML

Having recognized the value of the mainframe, the key issue becomes extending the data without disrupting the reliable operation of the mission-critical applications. Incorporating native XML logic into mainframe COBOL programs achieves this. The programs are converted to n-tier architecture and a new XML connection to the presentation layer is created. Tailored XML parsing logic is inserted into the program and a pre-compiler is used to translate the XML logic into executable code. The only changes to the program are the addition of a parsing module for the XML encoded data and the addition of logic to call the module. The business logic, I/O logic, and even the CICS logic in the program remain untouched. The data is encoded with XML tags as defined in a schema and passed back to the wireless platform.

Native XML bridges the gaps between XML developers and mainframe developers - without adding the extra maintenance and cost burdens of screen scraping. The same COBOL program running under CICS can output to browsers, handheld devices or 3270 terminals. The program's CICS logic is used if the input is from a character-based device and XML encoding/decoding logic is used otherwise.

First, the cost savings, when compared to the cost of purchasing and maintaining screen-scraping software, are significant. The incorporation of native XML, from a maintenance perspective, is no more complex than the addition of an EXEC CICS function. Further, when a change is needed, it is made at only one point.

Second, no middleware is required between the application and the web server. XML-modified programs become free-standing components that can serve as building blocks for new applications, while readily communicating with other programs and a wide variety of devices - from green screens to handheld devices. The advantage of this is that the data continues to be maintained by the trusted code that previously maintained it.

Delivering XML to the Device

To meet the challenge of delivering data to a handheld device a strategy must be developed that permits:

• Access via multiple devices
• Unique presentation logic for small screens
• Tightly-coupled, bi-directional access to existing back-end enterprise systems
• The combination of information from multiple sources
• Full operability even when the user is out of wireless network coverage.

This architecture supports access to multiple devices by making use of a Device Independent Rendering Layer (DIRL). The DIRL automatically renders an abstract description of a page from XML into the suitable markup language for the target device. The DIRL is equipped with an automated detection system that recognizes the specific device making a request. In addition, a user interface library delivers the exact representation appropriate for the particular device.

Equally important however is the ability to develop unique user interface logic that understands the limitations of the smaller screens on mobile devices. This architecture allows for custom tuning of output for particular devices. Whereas the DIRL will do the job of presenting the same information on a variety of devices, for many applications precise tuning of the presentation to the parameters of a particular device is very important. In these cases, by using a mechanism similar to class overrides, it is possible to specify a slightly different output for a particular device while the bulk of the application remains common across devices.

The task of achieving tightly coupled, bi-directional access to existing back-end systems is accomplished by leveraging the XML-based architecture of this solution. XML is used at the application level - meaning that specific business process operations (such as "place order" or "get customer ID") are represented by a library of XML-based commands. As described in the previous section, there exists a reliable and cost-effective means of generating XML connections to mainframe applications. This architecture leverages that approach to its fullest extent by applying business process orientation to those connections. For example, a single XML command such as "GET ORDER HISTORY of CUST ID 123" can be used to generate the necessary operations in XML to retrieve that data using multiple CICS transactions, if necessary, and even multiple systems. With each operation now operating as an XML stream, the retrieved data of ORDERS and the subcomponents of each ORDER (DATE, LINE ITEMS, SHIP STATUS, TOTAL COST, etc.) are also seamlessly represented in XML.

This approach has the significant benefit of giving the DIRL the type of data it needs to represent the proper workflow to the end user. In addition, there can be a myriad of disparate systems from which the information is retrieved because the XML "objects" which represent ORDERS (or other mainframe-based business objects) are independent of the XML branch commands used to retrieve the data.

Mainframe to Handheld Architecture

From a software perspective, the architecture now looks like the diagram below. Note that the only modifications to the mainframe applications have been the insertion of the XML parser logic and the separation of the presentation layer. The software for the web server consists of standard objects that are customized to fit specific devices and presentations.

So a solution to the problem of extending mainframe applications into the world of handheld devices exists. The architecture and software is on the shelf and available. It works. The final question is whether it is right for your application.

Determining the Right Applications

Mobile devices, from cellular phones to laptops, all have one thing in common - they have been developed to improve the productivity of field personnel. The obvious value of this strategy is in applications that support, either directly or indirectly, a work force that is based in the field at least part of the time.

Consider the following potential areas:

Mobile Sales Force
Enable sales and customer service representatives to directly capture information while meeting with the customer, place orders for immediate attention, or gain access to product catalogs, customer pricing and order history. The result: increased employee utilization, increased customer satisfaction, increased profits, and reduced cost of sales.

Mobile Field Service
Allow field service personnel to respond to and resolve issues as quickly as possible and to focus on issue resolution, instead of administrative tasks and travel time. The result: increased employee utilization, decreased time to close a ticket, enhanced customer service, and reduced call center costs.

Mobile Retail Management
Enable representatives to capture inventory audit information, access order status information, order new products, inform store managers about new promotions, reduce stock delays, and prevent lost sales opportunities. The result: increased revenue, increased sales, increased representative efficiency, ensured competitive advantage, guaranteed quality assurance, and secured compliance with company promotions.

Mobile Claims Solution
Allow claims adjusters to capture information during inspection, verify coverage and submit their estimates immediately for processing. The result: increased productivity of adjusters, improved customer satisfaction, reduced liability costs, and overall reduction in the costs associated with processing claims.

There are a number of other potential applications - wherever mobility and productivity are key business drivers. The crucial question to ask is whether a business function would be better served by IT if the information could travel with the person, rather than the person traveling to a terminal.

(c) 2001 Forecross Corporation

Forecross is a registered trademark of Forecross Corporation.
Copyright © 1996-2008 Forecross Corporation
All Rights Reserved.