Advice For the IBM i Application Fence Sitters
March 5, 2018 Dan Burger
Most executive management does not recognize the importance of IT. They don’t understand it and, therefore, don’t see the potential business advantages that come with restructuring a system that was put in place 10, 20, or even 30 years ago. So, how important is it to influence the perception of IT within organizations?
And how do you get more business managers thinking strategically about IT? One of the keys is to demonstrate how IT is the engine for growth and not a money pit with high risk and little return on investment. There are challenges to overcome, for sure. Don’t sweep them under the rug.
All this came to mind as I was listening to a webcast that endeavored to explain how IBM i Web and mobile applications can help grow business. During the presentation, the audience was asked about their plans pertaining to their existing core applications. More than half responded they were still trying to figure out what to do. Presumably they were looking for ideas and that’s why they tuned in to this webcast.
It’s not going to be easy to convince executive management that an IT plan, much less a strategic direction, is a worthwhile investment when you’re still figuring out how a project to modernize core business applications can be accomplished.
As one of the webcast speakers – Greg Patterson, an IBM i modernization expert at Fresche Solutions – said: “Dreams won’t come true without funding. Lack of funding or underfunding will not bring success.” Details that deliver a business impact need to be developed. A few examples of the types of projects that have significant business impact are: the capability to open new markets; the development of a self-service, e-commerce website; or the integration with other applications – a two-way street for information access. The flavors of complexity range from vanilla to rocky road.
Patterson’s advice begins with providing executive management with an assessment of the current state of IT. There needs to be a roadmap that includes an accounting of skills, resource availability, risks, costs, and scheduling, Patterson says.
Skills will most likely need to be acquired. Deploying applications to the Web requires a shift from developing apps to an emulator. New languages and new architectures are involved. You either train to obtain the skills, hire new employees with the skills that are lacking, or contract with professionals who have been down this road a few times before.
Fears of a skills shortage are greatly overblown, in my opinion. Successful organizations are training their existing RPG developers – the guys with all the business logic knowledge – and hiring young non-IBM i developers who pick up modern RPG skills quickly. But modernizing and modularizing applications must be part of the IT assessment. Admitting that some apps will require rewriting and that outside help might be needed to complete the assessment must be factored into the plan.
Core applications in IBM i shops are almost always heavily customized green-screen apps. They were originally written for internal use and for limited integration with other platforms. Unchanged, their business value is often questioned when compared to applications that are used by customers and business partners and are enabled for mobile users. To maximize business value, it becomes a modernize or migrate decision. Both options should be considered. One side or the other doesn’t always prove to be right or wrong.
A good decision is based on knowledge. Executives are known to demand more and better information, so they can make better decisions. Yet, from stories I’ve been told, the decision on what to do with legacy applications often gets made with insufficient information, at least regarding the options from the IBM i side.
The choice of options depends on the business needs and the IT strategy for the next year, three years, or even five years. They include:
Refacing existing green-screen applications by Web enabling them and incorporating tabs and drop-down menus to enhance navigation. This is the quickest, easiest and least expensive option. It’s also a very popular option. Minimal training is required.
Extending existing applications by incorporating new Web apps or Web services that call the existing core logic. This is quicker, easier, and less expensive than a total rewrite. Some training is possibly required to familiarize existing staff with the concepts of Web development, or these skills can be acquired on an as-needed basis from consultants.
In some instances, a complete rewrite is the best choice. Sometimes monolithic RPG code is so convoluted that it’s simpler to rewrite than make sense of the application so that pieces could be reused. Sometimes an organization will decide to rewrite RPG applications in Java or another language (PHP or Node.js, for instance) that provides platform independence.
Another option is to replace the existing application with a packaged app. This may or may not allow the use of existing business logic. When that’s possible, it should lower the cost of the project, allow faster project completion, and lower the risk of getting an application that doesn’t deliver the features and functionality of the original app. When little or none of the business logic can be reused with a packaged application, the risk of losing features and functionality is much higher and project failures are common. For all the time and money poured into these rip-and-replace options, the result can be a fraction of the previous code coverage and poor performance.
Well-planned and properly resourced projects increase the chance of success whether it’s migration or modernization, but leaving the platform has not worked out well for many migrations.
On the topic of well-planned projects, Patterson recommended the use of agile development processes, which allow a focus on smaller portions of the overall project and getting incremental feedback from the business users before continuing to the next phase. Engage with the users’ pain points, workflows, make them stake holders and a part of the process.
He also suggested that project planners think outside the IBM i. Consider access to data from other sources: Web services, other platforms, and other databases probably need to be integrated. Newer technologies and sources of information will almost always enhance the functionality of the application.
Deployment is not to an emulator anymore. Applications can be hosted on client side or server side. Train existing people — they have a wealth of knowledge about the existing application. Do they have web dev skills? Need the folks with RPG and database skill sets.
Without the details and a value proposition, the likelihood of executive sponsorship becomes very slim.
The webcast I listened to is available on-demand at this link. It includes several examples how organizations have successfully modernized applications to align with business strategy.