Application Modernization Priorities Meet Simple Solutions

August 16, 2011 Dan Burger

Complex problems are not always solved by complex solutions. To think otherwise is setting a trap for ourselves. Application modernization is a great example. Most times it’s not as intimidating as it’s made out to be. For instance, Valley Power Systems, a $180 million distribution company, was handcuffed with a business to business gateway that relied on Telnet sessions. Its modernization project, using looksoftware, took off the cuffs and took advantage of the Internet, smartphones and tablet devices.

Valley Power Systems is a dealer and distributor for engine, transmission, and equipment manufacturers. Its customers are in industries such as trucking, transit, construction, manufacturing, agriculture, mining, marine, power generation, and firefighting equipment. For more than 60 years, it has supplied complete engines, component parts, and provided repair services. Facilities are located in seven western states and the Baja region of Mexico.

The IT department supports direct order entry, inventory, warranties, and corporate accounting functions using an IBM Power Systems Model 520 and a mostly home-grown, custom software system designed specifically for its distribution business.

In 2010, Valley Power executives made the decision to replace its (I dare say decrepit) customer access gateway, through which dealers looked up products, prices, and availability. It was slow; it was expensive; and it depended on green-screen presentation of data, which some customers liked but a growing number of other customers did not appreciate and complained that it was an obstacle to productivity.

So as that quest for a new gateway played out, the debate about what to do with the green screens began. As is the case with many home-grown applications in companies with some history behind them, there is a considerable amount of very specific data being collected. To rewrite applications with similar capabilities would be a monumental task. And to discard the existing apps would be flushing money down the toilet.

Naturally there was a deadline. In this case, it was January 1, 2011.

The pressure was applied when the gateway provider determined it would no longer offer that service. So Valley Power knew what it had to get done. The goal was having a Web-based customer gateway–a portal–and customer-facing applications with a graphical user interface in place. The clock was ticking.

Rich Merchak, IT manager at Valley Power, researched several strategies. One of those was quick and dirty screen scraping–giving the 5250 a very simple graphical interface. Merchak recognized this was not a good long-term solution to the problem. Eventually more application functionality would be required, new application development would come into play, and the screen scraping plan would become a dead end. Yet, with little time before the deadline there was not much chance or learning a new development language and undertaking an application rewrite, even if the goal could be reached by modifying only five or six of the highest prioritized applications. Time was not on their side.

It was already late in the year when Merchak connected with looksoftware.

The deadline was quickly approaching, so a looksoftware consultant was brought in to make the best use of a limited amount of time and make certain the required applications were up and running and capable of handling the workloads.

The highest priority involved screens that didn’t require a lot of massaging. It didn’t require combining fields from multiple screens or tabs or changing the ways users navigated through the system.

“The initial focus was on the customer,” Merchak says. “It required four or five screens being enabled. The consultant built it out and showed us (the IT staff) what he was doing and how he was doing it. We learned by example.”

Watching is one thing; doing is another. The looksoftware consultant was only on-site three days, and then Merchak and programmer Tad Peterson got in the game.

“After a year of reading articles about PHP and .NET, and trying my hand at HTML, it surprised me when I saw what I could do without reading a book or going to class,” Merchak says. “Not everything came easy; some things require thought, and there’s some trial and error. Finding our way didn’t take long though. Once we learned a few things, we were even cleaning up our old applications because we knew what it was going to take to work in the GUI with the buttons and drop-down menus. If you put the right options and the function keys in the right place, the buttons and pull downs will work without a hitch and we never have to write any HTML.

“It was a matter of seeing what was not working properly and being able to quickly find the reason for it and applying the fix. When we had questions that we couldn’t answer, looksoftware provided expert assistance.

“And then we began to see what could be done outside of the RPG programs by using looksoftware. Unwanted text can be blocked from appearing. There’s no need for numbers next to everything like you see on the green screen. I think we are pretty proficient now with this as far as we’ve taken it.”

As an example, Merchak said a few weeks ago there was a problem with a mobile app running on the iPhone. The original iPhone apps were created months earlier and he wondered whether he would remember his way around that app. Not a problem, he says, as he was able to navigate through it, find the problem, and fix it.

The transition to the new Internet portal has gone well. It’s been a gradual process. Rather than just pulling the plug on the old system on January 1, there was a gradual introduction to the portal. Some customers were already using a link on the Valley Power website to log on to the Telnet sessions in the virtual private network where business was transacted. On January 1, that link took customers to the new Web-based pages.

“I was expecting to get bombarded with phone calls immediately after taking down the old system,” Merchak says. “Amazingly the old customers who used the green screen every day never called. I thought everyone would say ‘What do I do now?’ Because it was all so simple, they quickly figured out where to go.”

End users do have the capability to toggle back and forth and view applications in green screen or GUI, whichever is their preference. Use of the new gateway continues to increase.

Customers that clung to the old Telnet gateway were rounded up and converted gradually over a period of four or five months. This process allowed the new gateway to be closely monitored and software licenses to be adjusted according to usage. The licensing is based on concurrent users and Valley Power currently has 30 licenses to handle the traffic.

Since replacing the old gateway, Merchak estimates Valley Power is saving between $10,000 and $15,000 a year in fees for that service. He also points out that the company saved a lot more money by being able to reuse rather than rewrite the existing apps.

Companies that remain on the fence about application modernization can gain some perspective by seeing what Valley Power has accomplished without getting mired in complex time- and labor-intensive projects that are budget busters. It took a phased approach and put as much effort as it could into accomplishing a few priorities.

And IT managers programmers who are limited by the amount of time it takes to learn new skills can also come away from this with a lesson.

“It’s like a lot of things in this industry,” Peterson, the programmer at Valley Power, says. “We went to school years ago and learned certain things. A lot of what we learned, we don’t do any more. But what we really learned in school was how to learn how to learn.

“It applies to what we’ve been talking about today. By sitting down and pushing that application into production in less than six weeks, it gave us on-the-spot, intense use of the system. It was probably the best training we could have gotten because we did the same thing over and over while making sure it was working for us. That’s how you remember steps later on. It’s not some theory you learned in class. This involved applying things and seeing how it was working.”