Open Source RPG Apps: The ‘Bright Future’ That Didn’t Happen
July 10, 2006 Mary Lou Roberts
Flashback to 2002. Open source software is hot. It’s the future. It’s what everyone is writing and talking about. Projects and their related Web sites are popping up all over. And it wasn’t just the Linux and PC folks doing the talking either. There were all kinds of open source projects for systems and application software–tens of thousands of projects, many of which called SourceForge home.
With apologies to our own Alex Woodie (and with hope that he won’t retaliate by digging up some of my headlines four years hence), IT Jungle published an article entitled Open Source is Alive and Well in OS/400 Land. In the first paragraph, Woodie said that, “Despite its proprietary nature and reclusive status, the iSeries has a bright future when it comes to open-source application development.” While he immediately acknowledges that much of this development focuses on Linux and PASE, the OS/400 AIX runtime environment, he goes on to note the open source application development taking place in RPG as well as Java and C.
To be sure, there is still a cadre (though far from a treasure chest) of open source software available to iSeries users written in languages other than RPG. But what ever happened to open source RPG applications? Most of the Web sites that hosted open source RPG projects have either disappeared or have not been updated in years.
John Ross, president of JMR Consulting, which runs the NetShare400 time sharing site, points to a few Web sites and tools that are still alive, including some projects on SourceForge, one called Uzaemon, as well as the apparently now defunct project that many thought would be big, WyattERP. (This was an open source ERP project written in RPG, and it has the best name of any project, ever.)
Ross is pessimistic about the future of open source RPG applications. “Since most people do not have an iSeries of their own, the company they work for has to allow them to work on the projects on the company’s computer, and most companies I have worked for have rules about this and who owns the code. This makes it hard to do open source projects. I also don’t find many RPG programmers wanting to spend their free time on open source projects. I’m not sure if it’s our age (I figure most are older).”
Scott Klement, IT manager and senior programmer for Klements Sausage, has been involved with open source RPG projects for years, and is the developer of several tools, which he makes available on a personal Web site. The most widely used of these are HTTP API, an RPG service program that uses socket calls to implement the HTTP.1.1 protocol to transfer documents over the Web, and FTP API, which uses socket calls to implement an FTP protocol, reports errors if anything goes wrong, and gives the program the ability to retrieve the error message.
Klement maintains that approximately 1,000 people are currently using these programs (though only about 20 actually make a contribution in helping with development) and that use seems to be increasing, although he doesn’t track statistics carefully “because it isn’t a business for me.” What types of companies use these programs? “The smart ones,” quips Klement, although he sees no other profiles in terms of company size or industry.
A strong believer in both open source software and RPG, Klement has been in a unique position to build and offer utilities such as these for the open source community, since he works for a family business and has access to the company iSeries from home. (Klement Sausage was founded by his grandfather and his two great uncles.) He does this work because he enjoys it and because “it helps make the products that my company uses better.” And, because he offers the software to others, they help to find bugs and report them and help in fixing them, and it works in more environments because it’s being tested by people in different areas. “That helps make it better software that will work better when I need it myself,” he says.
However, Klement doesn’t mask disappointment that open source applications written in RPG have not been more widespread. He attributes the lack of availability of open source RPG programs to the fact that “most people don’t have an iSeries that they can access from home in their spare time. Open source doesn’t work out too well on the iSeries because most businesses don’t want to devote their resources to developing anything that doesn’t directly help their bottom line.” Typically, he points out, Linux or PC programmers have the ability to work at home.
Looking forward, Klement is hopeful but not overly optimistic about the future of open source RPG programming. Although he believes that people are warming up to the idea, he questions whether or not people will have the resources and the desire to do it. “It used to be that people were very skeptical of the whole idea. That’s not the case anymore. Now even the most conservative people in the iSeries world understand the value of open source. It’s a question of whether people can find a way to work on these projects and have access to an iSeries from home.”
The irony, of course, is that shared source but paid-for licenses for software were the norm for RPG applications for decades in the IBM midrange market. The SSP, CPF, and OS/400 communities were so close to going all the way to open source, long before Linus Torvalds even thought of making a clone Unix environment, and one of the key differentiators of RPG applications was that customers got the source code and could make their own changes so those apps better fit their business. So close, and yet so far.
Of course, for open source RPG applications to be created, RPG itself must survive, which Klement sincerely hopes it will. While he acknowledges that the aging RPG programmer base is a concern, he sees company commitment to RPG as a bigger issue. “We need to reach people and get them to upgrade their skills and learn how to write RPG software in a manner that’s more resilient and easier to work with–to start making Web interfaces for their programs and to use SOA and other new concepts.”
If open source software written in RPG is scarce, that doesn’t mean that the RPG community doesn’t have access to a wider range of open source tools. There are several tools written in Java that target the iSeries audience.
Steve Gapp, president of SoftLanding Systems, agrees that the most successful open source projects supporting RPG development tend to be written in Java or other languages, and they tend to be utilities rather than applications. “This is borne out,” he says, “by the activity we see in the two open source projects we host, RSE Extensions and Subversion for OS/400. For example, the RSE Extensions have been updated twice in less than a year, and they’ve seen about 1,300 downloads in that same time frame. We see similar activity for Subversion for OS/400.”
In Gapp’s experience, RPG programmers use open source tools mainly because of their adoption of WebSphere Development Studio and their application modernization efforts. Other than that, like Klement, he sees no patterns in company size or industry. “If it [an open source solution] solves a problem for them and if it is reliable, customers will use it. However, that usage does tend to be a one-way street. We don’t see too much in the way of contributions back to our projects.”
As for the future of open source RPG applications, Gapp is not hopeful. “I do not see too much in the way of RPG in the open source world because most RPG code out there really supports heavy-weight business applications. It’s not likely to be used as the development language for open source projects–even those associated with OS/400.”
Martin Rowe, senior analyst/programmer with Moores, a large furniture manufacturer based in the United Kingdom, offers his database generation utilities, dubbed DBG400, to the OS/400 community for free. (“Free” is not the same as “open,” but I’ll get to that in a moment.) He comments that he has not seen many large initiatives, other than WyattERP, and he hasn’t heard anything about that project in years. “My impression is that most of what’s available is similar to my own offerings: small utilities designed to scratch a developer’s particular itch. These can then turn out to be useful to others. Most of these are programmer tools. I can’t think of any that are designed to be used directly by end users.”
Rowe comments that many people use his utilities and others like them not necessarily because they fill a need that cannot be met by software on the market, but because management may not sign off on the purchase.
Bradley Stone, president of BVS Tools, agrees. “I’ve tried open source a few times and it was a failure in my opinion. People treated it like free software instead of a group/community effort, so I ended up supporting a free piece of software and the code that they had played around with.”
And, indeed, it does seem that what drives open source RPG programs (to the extent that they are driven at all) is the “free-ness” rather than project collaboration. Now, if one wants to talk about “free” software for the iSeries, one need look no further than IBM for tools like Toolbox for Java (the open source version of IBM’s Toolbox for Java), a library of Java classes that allow Java programs to access i5/OS data and resources.
However, as Simon Phipps, who runs open source software at Sun Microsystems pointed out in his 2002 article, Free Speech, Free Beer and Free Software, “‘Free’ is used as in the phrase ‘free speech’ (a right we cover), rather than the phrase ‘free beer’ (always too good to be true) or ‘free kitten’ (which sounds good, but has a high overhead) . . . . The early years of open source . . . focused on free (as in beer) software, so it is still possible to misunderstand. But we have seen a definite shift in thinking. The open source community has welcomed companies that build commercial enterprises, as long as they act symbiotically rather than parasitically.”
Phipps calls the open source community model a “community of code,” which has a gatekeeper who embodied the will of the community to set bounds for the project and gives the community stability. Looked at this way, by almost any measure, one would have to conclude that open source RPG software has been a failure–if it ever really existed at all. Yes, there are pockets of small successes such as Klement and his 20 or so contributors. But is that a community of code? No.
“Of the various projects underway at the time I got started, not many are still actively developed,” Rowe comments. “Some people took their projects commercial. Other changed jobs or moved away from the iSeries. Others no longer had the time available to work on their projects. I keep my Web site updated, but unfortunately no longer have the time at work to develop anything new unless it fits in with a work project. That’s true of any open source project, though especially so in the iSeries world, given the business nature of the platform. While it’s possible to get a timeshare account on an iSeries, that’s not the same as having your own home PC to develop on.”
Most of the successful open source development efforts that will benefit iSeries users will probably focus on Java, giving the developers a much larger pool of contributors and resources.