Advice From The Mavens Of Modernization
June 29, 2015 Dan Burger
You don’t have to stretch your IBM midrange imagination too far to understand that today’s app dev decisions will have a great impact on tomorrow’s IT success. Some things never change. And that’s one of them. But situations change and doing business changes. The trick is to integrate the past–keep the valuable experiences and toss the excess baggage–with the present to prepare for the future.
Technology is almost always disruptive. How you deal with that, how you make it work for your career success and the success of your organization is the on-going challenge until you decide to hang up your spurs.
A lot of organizations that have been successful in the past (and the present, I might add) owe a measure of gratitude to the IBM midrange computing system. For the most part, its service has been exemplary. And for the most part, users’ decisions have been reasonably sound. But for many companies, what got them through the past 20 years will not get them through the next 20. There are some changes that need to be made.
I asked a few IBM i old timers what they think about being prepared for what’s ahead.
How will application development change in the next five years?
Roger Pence: AD, in the IBM i midrange, will become more collaborative, less monolithic, and more focused. We’ll spend much less time building “the” app, and much more time building services and domain models–the building blocks needed to build future apps. “The” app goes away and smaller, more nimble mobile apps consume the services and domain models.
Duncan Kenzie: On the IBM i, I see a continued move from green screens to Web applications; the platform is still not where it should be in that regard. Those modernized screens will do a better job of integrating data from multiple systems, using AJAX and JSON to do that and move away from XML. There will also be more of an emphasis on mobile development which is still not as widespread as you would hope on IBM i. More organizations will use PHP or other open source languages with one reason being to attract younger programmers. This will lead to more rapid and automated development tools, more server-side and client-side frameworks. In the non-IBM i world I expect programmers will move more to single language development, like Node.js, which can handle most of both the client- and server-side code.
Steve Gapp: Did anyone predict the business impact of the iPad when it came out five years ago? The only five-year prediction we can make with any kind of certainty is that today’s hot languages and frameworks will be passè and there will be another unforeseen, game-changing technology that will catch us off guard. That said, IT must constantly strive toward’ being agile enough to react to these changes in a timely manner.
Chris Koppe: In the IBM i space, I expect that more developers will adopt tools that can help them work faster and smarter–tools with rich development environments–as well as application analysis tools. With the on-going adoption of (and transformation to) Web and mobile, developers will need to learn and master these new interfaces and the tools used to build them.
What advantages or disadvantages will companies that are looking five years in advance experience compared to companies that are taking an approach of solving the problem that is in front of me today?
Steve Gapp: Companies set goals and create objectives to ensure growth and stability. But focusing too much on short-term, tactical initiatives rarely addresses these issues. Instead, think long-term and use IT for the betterment of the business as opposed to it being a constraining factor.
Roger Pence: You snooze, you lose. There are still IBM i shops today that think it’s OK to wrap 75 percent of the app around DFU. There are still shops today that think Web and mobile are fads. Forward-thinking companies today are looking to improve workflows and increase competitive merit with new skills and new ideas. There are zero advantages to avoid looking ahead. Decision makers who continue to think that RPG is the best thing since sliced bread will be the first in the bread line.
Duncan Kenzie: Companies looking five years ahead will remain competitive. They’ll be able to take advantage of modern technologies, of mobile devices, of the efficiencies and usefulness of integrating the endless data from other sources that is now available. That data includes social media tools to collaborate–including those tools in ERP applications to engage employees–and sensors from the Internet of Things. Companies not looking ahead will lose ground to competitors, will end up with a retired workforce, and will not be able to attract younger programmers. One day they’ll have to leave this great IBM i platform at a big cost and risk.
IBM executives like to talk about integrating systems of record (SoR) with systems of engagement (SoE). How do you see this taking place?
Roger Pence: Oh, Lordy, do I hate couching evolution in freshly minted fancy talk! We’re all stuck with huge, lumbering apps that provide the business’ core IT backbone. An IT director should always spend a portion of the day keeping the system oiled and running, but also needs to be spending the majority of the day planning, and building, incremental, supplemental apps to bridge the capabilities of legacy system with the demands for the future. These new apps are mobile, focused, and impose dramatically new and improved workflows. We’ll start with clusters of small apps, but ultimately they will own our world. The conundrum is that no shop can to afford to rip and replace its legacy system, nor can any shop avoid the reality that its workhorse app must ultimately go away.
Duncan Kenzie: The technologies for integration are already available using JSON, Ajax, and RESTful services. The challenge is for IBM i IT managers or CIOs to make the SoE business case, or to at least take the first steps of creating Web applications and integrating data from other servers. Infor’s done a good job of this with their SOHO UI and Ming.le. One big challenge for IBM i shops is the UI design; you need a UX designer to help get the most from the applications.
What are the integration issues right now and how might this change in five years?
Steve Gapp: The biggest integration issue today is SoR’s lack of openness. In order for SoE to be successful, seamless integration is required. Over the next five years, companies will continue to open up the SoR with more and more APIs of varying types in order to enable core business functionality to be consumed by any type of SoE. Businesses can no longer be constrained by legacy applications if they wish to remain competitive in the future.
Roger Pence: The biggest issue today is how can IT continue to offer the services that fulfill your business’ unique competitive advantages, while at the same time be rationally decommissioning parts of that app and substituting those parts with nimble, modern mobile replacements? Our community, bless our hearts, moves slowly. The advent of mobile, and the pressure it places on competitive advantages, will be a disruptive change like we’ve never experienced in the midrange. The gravy train for the IBM i decision maker afraid to make decisions derails within the next five years.
Chris Koppe: Integration with Web and mobile technologies and integration with other discreet, non-IBMi applications are among the integration issues. And integration with new interfaces and other applications will continue to be a pressing issue. Many new applications will come with sophisticated Web interfaces, and integration will require strong Web and mobile development skills.
Will there be pressure to consolidate SoE and SoR systems into a single system to simplify infrastructure in a server consolidation type of strategy?
Duncan Kenzie: There will be more pressure on IBM i shops to integrate different data into existing SoR systems. This will provide a fuller picture of the information people need to do their jobs more efficiently and effectively. There may be pressure, but to a lesser extent, to add the collaboration features to the SoR applications. Ideally they can modernize or transform their SoR to a SoE rather than them being separate entities.
Steve Gapp: Yes, especially when companies need to integrate big data and core business applications to provide business analytics and key performance indicators to executives and line of business managers, which in turn enables companies to make better business decisions.
Pascal Polverini: It is not that both systems will become the same, but they are two ends of the same rope. On one end you have processes and on the other end you have people. SoR will pull the process side and SoE will pull the people side. The process side develops more analysis and structure (relations, BI) and the other side more UI and accesses (mobile, real time, geolocation, etc.). On one end, the system is the context. And on the other end, the context is the system. More harmony needs to be defined between business, private, and public data, but it is undeniable that the rope will become shorter. Technically, SoR will have to embrace SoE to enable intelligent processes, but from the front-end it will look like SoE and SoR are the same.
Roger Pence: Let’s dispense with the Gartnerspeak: SOR = old school and SOE = new school. Of course there will be pressure. The biggest pressure possible.
Will the cloud play a significantly different role in five years in terms of integrating SoR and SoE? Will we see a lot more infrastructure as a service and platform as a service?
Steve Gapp: Over the last five years companies have become increasingly comfortable with the cloud and have already shifted certain business segments off-premise and into the cloud–email, CRM, HR, and more. There is no doubt that over the next five years companies will continue to shift more workloads and applications from behind the firewall to the cloud.
Duncan Kenzie: Most IBM i shops will not want their core business data running in a public cloud, but they would benefit from integrating their SoR systems with cloud-based applications they use such as Salesforce. I would expect much more integration like that to happen over the next five years, as well as more virtualization and organizations running private clouds within their firewall.
Chris Koppe: Definitely. It may start with development servers, but eventually many companies will look to the cloud for computing platforms and power. This will be carefully assessed, as there is (and will continue to be) a need to feel in control for quite some time. Although the cloud is very safe, there will be reluctance by old school IT leadership to adopt the cloud. There are also economic reasons why the cloud does not make sense for everyone. It is almost never cheaper than traditional on-site computing.
Roger Pence: Absolutely. Services, clouds, collaboration, teamwork, integration, soft skills (fancy talk for recognizing you are no longer coding by yourself), and DevOps are in the future for all of us. For those that disagree, a hamburger flipper awaits. In fact, in honor of IBM’s infamous hamburger flipper ad of 20 years ago, let’s all just agree to label the upcoming IBM i application revolution HF 2.0. (In 2060, grandkids will ask grandparents, “How did you survive the HF 2.0 revolution?”)