What IBM Can Learn From Free-Form RPG
August 31, 2015 Dan Burger
Three weeks ago, I wrote an article about free-form RPG, a runner-up in the IBM best kept secrets game and second only to magnificently camouflaged IBM i. Since then, I’ve collected the opinions of several more free-form observers. I’ve not talked with anyone who believes IBM is doing enough to encourage free-form conversion of old RPG code. But there is more to say about whether the benefits and the effort it takes to make the code conversion.
Technically, the conversion from fixed-format to free-form RPG is pretty simple. Tools from IBM i ISVs ARCAD and Linoma Software are capable and available, though like free-form itself, are not marketed with the vigor of mobile computing, data analytics, or even social media.
I hear all the time that the IBM i community is slow to change. Does that explain the minimal efforts in raising awareness? Do even the best efforts at raising awareness fail? The lack of awareness issue was the point of my first article. I’m moving on.
Some of what I gathered since the first article took the awareness point to a different place. It’s not that the IBM i community is totally unaware of free format RPG. It’s that their awareness is based on the old free-format RPG, which was much more limited in what it could do.
When free-format RPG was introduced, its inclusion resulted in code that repeatedly skipped back and forth from fixed-formatted to free-formatted and back again. It may have been designed as a first step toward free-form RPG, but it just may have sent the message that free-form is more trouble than it’s worth, which is likely because it really wasn’t worth a lot. People still think free format has the problems it had when it was first released.
The original free-format RPG made existing code worse, in the opinion of Phil McCullough, the director of information technology at the American Registry of Radiologic Technologists.
At the Registry, the IBM i-centric staff is converting the mostly RPG-LE code to free format. The reasons McCullough gave relate to unifying the multi-platform develop environments. “The often-heard description that RPG developers don’t talk with the .NET developers and vice versa gets broken down a bit with free form,” McCullough says. “We want to bring the RPG and .NET development groups together and get the RPG developers out of their proprietary and isolated world. Along with this comes the use of RDi, which puts RPG developers in the same ball game as the other development teams.”
Like many IBM i shops, there are .NET programs accessing data stored in DB2 for i. The developers can do this without interpreting the RPG code, but it becomes easier and more efficient when .NET programmers can make sense out of the RPG code and the RPG programmers can better understand .NET code. Free-format moves in that direction.
“We will eventually convert 100 percent of our RPG to free form,” McCullough says. “We can’t just stop and do it all at one time. The goal is to do this over the course of the next year.” The conversion process is being accomplished with the help the RPG Toolbox, a plug-in to RDi. The Toolbox is a product of Linoma Software, which identifies the sales of its RPG Toolbox as steady, but declined to provide any statistics.
I got a similar response from ARCAD Software, which makes a tool called Transformer that also converts fixed-format RPG-LE code to free format. The sales of ARCAD and Linoma tools (which seem to be lukewarm) are indicators that free-form RPG is not a high priority for many IBM i shops. All things in due time, I suppose. The modernization of application development in IBM i shops can make turtle racing seem like high speed action.
Because I took IBM to task for not doing more to promote free form RPG, I’ll also mention that neither Linoma nor ARCAD is doing much in this regard either. IBM tends to look to the ISV community to promote IBM i and all the hardware and software that goes with it. The ISVs tend to believe, like I do, that IBM should do something to promote IBM i on Power Systems and all the software that goes with it.
McCullough believes the pendulum is swinging in the direction of free-form adoption despite the recognition that it takes “a little getting used to” before programmers get that comfortable feeling of familiarity.
“User group meetings and webinar participation is bringing it more attention. It’s being talked about like it’s standard stuff. The people in the PDM world are slowly coming to the realization that the modern technology is the way to go,” he says. “IBM should have a bigger budget for this. It’s to their own benefit to help move users in this direction. But, to their credit, they have been hanging in there and not pulling the rug out from under any traditional users. They won’t abandon the shops that continue to program in green screen and old RPG.”
Free format and the RDi development tools go hand in hand. And what I’ve said about IBM not doing enough to promote free form RPG goes for RDi as well. RDi is another product that suffers from its early release deficiencies. It’s always been difficult to get green-screen developers to drop their guns and come out with their hands up. And those who first explored RDi territory found it inhospitable, so they retreated to what they knew. RDi has improved, but the past is not forgotten.
Those are the cards that are dealt to the evangelists like Charles Guarino, Jon Paris, and Susan Gantner.
Gantner has been telling me for some time that interest in RDi is on the rise after years of little interest whatsoever. She credits free-format RPG for some of this, but believes the productivity features in RDi are better reasons to move to RDi that the ability to write a completely free-format program. A few weeks ago, Gantner presented a webinar hosted by COMMON on the topic of RDi. It was the most attended COMMON webinar of the year, according to Manzoor Siddiqui, COMMON’s executive director.
Gantner also says attendance at the RDi HeadStart workshops that she’s presented at the past few RPG & DB2 Summit events have been well attended and sometimes sold out.
Some momentum is better than no momentum, but I remain skeptical of free-form RPG becoming widespread in the short term. Until the pace of organizations prioritizing the unification of their multi-platform development environments picks up, the adoption rate of free-form RPG will rise and fall with the economic tides. Other factors that block the use of free format were also pointed out. For instance, a shop that’s dependent on third-party software written in fixed format RPG will hang back and wait until their vendors move forward before they put this on their plates. And you can add the all-inclusive Whack a Programmer excuse that old RPG hacks just refuse to learn anything new. One person I talked with–Ken Killian–called these people “punch card holdouts.” Killian has used free-from RPG for 18 months and loves it.
“It took me a little while to get used to it, just like it took me a while to get use to ILE RPG after coding in RPG/400,” Killian says. “You just use it and get used to it.”
After my first article on this topic, Bob Cozzi, a well-known RPG author and often a vocal critic of IBM’s RPG roadmap, sent me an email expressing his point of view. He says free-format awareness isn’t the issue. It’s the lack of a GUI that is the issue. But because we are on the topic of free-form RPG, he had this to say: “The current ‘really free-format’ version of RPG is a bit confusing and not well designed. It is substantially confusing to non-RPG programmers such that it probably won’t catch on beyond the usual group of lecturers and article authors who pretend it has become the standard when it hasn’t.”
Cozzi contends that IBM “killed The 400” in the mid-1990s by deciding a native GUI wasn’t right for the (then) OS/400 platform. That’s another story for another day.