In last month's missive, I started a series of articles discussing the connectivity choices you have between Linux and the iSeries machines.

Specifically, I discussed the open source project "tn5250," a 5250 emulator. That article generated a few emails that, for the most part, directed me to various emulators. Some of the products mentioned were open source, some were freeware, and the rest were commercial products. All are worthy of mention, so I'm dedicating one more column to the topic of 5250 emulators. But, this time, I'll do so in the context of the open-source movement.

This month's column was inspired by the emails I mentioned, and, in particular, an email that I received from one individual who lambasted both the open-source movement (for providing free, market-killing alternatives to commercial wares) and me (for "... corrupting and leading people down the path of inefficiency and declining productivity").

I've been writing about Linux and open-source software for a few years, and this is the first time I have received a negative response to one of my articles. The rest have all been positive. And why not? I'm not writing about religion, politics, or abortion rights. I'm writing about something that most consider to be a good thing. So when this email arrived, I was somewhat taken aback. Had he posted his comments in Usenet instead of emailing them to me, he certainly would have been branded a troll, but I assume that his comments were sincere. At first, I considered replying directly, but after further consideration, I concluded that it might be better to address his comments in this article. That way, my reply might benefit others who are having trouble with the open-source concept. (And I'll benefit, because my message will have gone through a couple of editors before being released, thus keeping things civil.)

The gist of this gentleman's email message is that IBM, with its AS/400 Client Access Express (CA/400) product, has brought order and productivity to the chaos that was AS/400 (and presumably System/3x) connectivity. He goes on to list the various products that he "struggled with" over the years and opines that the connectivity issue has been made moot with the advent of CA/400. In his view, open source only complicates a now-simplified marketplace, leading to declining productivity. And if everyone went to open source software, the economy would collapse and people would starve because open source is the antithesis of a capitalist system. (I'm not kidding. Paraphrased, that's what he said.)

Okay. There are still plenty of third-party vendors willing to sell us connectivity products. So I'm not sure why he believes that the chaos he alleges is gone. Assuming we buy into his premise that before CA/400 there was chaos and now things are oh-so-much better, we have to ask the question: What is it about CA/400 that made things heaven on Earth?

I would argue that CA/400 brought standards to AS/400 access. Old-timers will remember the various products available in the early days of 5250 emulation. You could buy products not only from IBM (Workstation Emulation and PC Support/3x) but also from a plethora of vendors, all of whom seemed to have their own way of getting data back and forth from the PC and the System/3x, although some supported the PC Support APIs. It seems that the problem wasn't that the other products didn't work (I had lots of good luck with many vendors,) but that each vendor did things its own way, making large-scale deployment somewhat challenging in a mixed-vendor shop. In my opinion, the single biggest thing that made CA/400 useful is that it offered shared folder support so that DOS/Windows and OS/2 programs could store and share data on the AS/400, albeit in breathtakingly slow motion. Authors of PC software now could use the CA/400 data transfer programs and, if they wished, use the AS/400's storage for a repository all by using the familiar drive letter. That was slick!

As time has passed, the Client Access product has evolved. It still performs essentially the same tasks plus a whole lot more. But for the purposes of this discussion, it's not what it does that should interest us but rather how it does it.

Gone is the proprietary network protocol formally used to communicate with the iSeries. Since twinax is now all but dead, IBM uses the industry standard TCP/IP protocol to perform its magic, thus eliminating an impediment to network administrators integrating the iSeries into their TCP/IP network.

Gone is the proprietary driver IBM provided to give the PCs drive letter access to the iSeries. Instead, IBM chose to write a server (NetServer) on the iSeries that implements Microsoft's Server Message Block (SMB) protocols so that the PCs can access shared folders using the native methods provided by the Windows operating system. Personally, I think IBM would have been further ahead if it had adopted the Samba project's version of an SMB server instead of rolling its own, but I digress. Come to think of it, gone is shared folders. That old dog now has been replaced with the Integrated File System (IFS), which is much faster and much more elegant than the shared folders support it replaces.

Gone, too, are the proprietary data transfer programs that were once part of the Client Access Suite. Sure, they're still there in a different form and you can still use them. But you also can use the standard ODBC and JDBC drivers in your own software or from within most office suites, such as Lotus or Office. And if you really want the data in a PC file, you can use FTP as well.

So, as far as I can see, it's not Client Access that has brought order to the iSeries universe, but rather it is IBM's commitment to making the iSeries embrace existing standards. Client Access is now just another way to get there from here, not a means to an end.

I don't want to sound like I'm beating up on Client Access. It's a darned nice piece of software, with great functionality, that makes the iSeries play well in a Microsoft-based network. But for as clever as the product is, it suffers from one serious design flaw--it runs only on Microsoft Windows, a platform that I eschew. I wouldn't even care about CA/400 at all except that IBM has chosen to make it the only native method to configure and control some features of the iSeries, at least without my having to write my own software to use their APIs. Thus, I am forced to choose between writing my own software or having a Windows machine around just so I can configure my iSeries. To make matters even worse, I prefer using a green-screen to control and configure my iSeries, so much of the functionality and the pretty GUI designed into CA/400 is wasted on me anyway. When IBM finally makes good on its promise to rewrite CA/400 in Java, then I'll use it. Until then, I view it as a product that is still part of the problem of proprietary platforms.

Now let's put reality on hold for a moment and examine the second half of this guy's argument--that if everyone moves to open-source software, then the economy will collapse and people will go hungry. Imagine for a moment that everyone does dump their existing software and adopt open-source software instead. What do you think will happen? His view is that open-source equals free. That's a misconception. Whereas the software is, in most cases, free for the download, an entire market of services has been created to support it, from magazines and books to consulting services. So, while there always will be those who can manage an entire network single-handedly, most people need help installing, configuring, and maintaining their software and systems. That is true for commercial software, and it is true for open-source software. The only difference is what you pay for the software initially and the license (and, therefore, control) you have over the software during its lifetime.

Take Microsoft, for instance (please). How long do you think they would be in business if all they had for income was the pittance they charge PC manufacturers for the copies of MS Windows and Office that come pre-installed on PCs? Given that PC sales have been flat of late, I'd say not very long. Their big income comes from the support and consulting services they provide to users of their software, the technology they license to others, the certification programs they offer, and the ordained upgrades that users must actually buy to get the latest bug fixes. This list isn't all-inclusive, but you get the idea.

So how is open-source software different? Assuming that you download the software, your cost is roughly equivalent to a pre-loaded copy of Windows, so let's call that a wash.

The same services for consulting and support that you can buy for proprietary software are available for open-source software. As I mentioned in an earlier column, even IBM provides those. So that's a wash, too.

The first divergence is that you no longer need to pay for upgrades to get bug fixes. Upgrades are available for download just as they were for the current version. Nor are you forced to upgrade. I've seen studies that show 80 percent of people who use word processors use only 20 percent of their available features. So why upgrade? In the proprietary world, you do so to get the benefit of the latest security "enhancements," which are conveniently not available on your current platform. But in the open-source world, you upgrade only the packages that you need to when you want to. It's not all-or-nothing.

The second divergence is in the control that open-source software gives you. Anyone who has ever developed software for the Windows platform can tell you horror stories about software breaking when Microsoft released a new version of its flagship product. For that matter, even service upgrades can cause a vendor to go crazy. Just ask IBM how easy it has been to maintain its CA/400 software. The upshot of all of this is that anyone who has developed software for Windows has been at Microsoft's mercy. The DOJ case has demonstrated that. And it's not just Microsoft; any vendor who provides proprietary software can become your worst nightmare.

Contrast that with the open-source model, where you actually have the source code to a product in your hands. If you don't like how something is done, you are free to change it. You don't have to worry that some other company will cut your legs out from under you because you are no longer dependent on their good graces. Open-source puts you in the driver's seat.

As for making money, there's nothing stopping you from making a profit on your work. Contrary to popular belief, you're not required to release your product as open-source simply because it runs on an open-source operating system. So those who have avoided open-source software because of this fear now can sleep well at night with this knowledge.

For final proof of the viability of open-source software, I have only to point to Big Blue. You can't find a better example of capitalism than IBM, and IBM doesn't seem to be intimidated by the open-source model. Indeed, IBM has dumped huge bucks into open-source software development, and I don't think it did it out of the goodness of its corporate heart. It has determined that open-source fits into its business plans. It can (and has) made money from it. And IBM isn't the only one. It seems everyone is jumping onto the open-source bandwagon in one way or another.

I could go on for a long time about the advantages of open-source software. But there are already plenty of treatises available on the 'net and in the bookstores. Instead, let's now turn our attention to the topic of 5250 emulators. I'm not going to do a product review. Instead, I want to discuss what open-source has done to the marketplace and scope out the existing landscape. The short list of products that I'll mention is certainly not all-inclusive. It is meant to be representative of what's available on the market. I apologize in advance to the vendors of the many products I'll fail to mention. For a more comprehensive list, please check MCPressOnline's oasis of software and hardware information, also known as the Vendor Directory.

A discussion of license terms during the early days of 5250 emulation would be senseless, since, for the most part, you'd use the software that came with the emulation card you installed into your personal computer. Some manufacturers provided updates to the software for free while others charged for anything more than simple bug fixes. There were some products, such as NetManage's Rumba, that were capable of running on top of IBM hardware. But usually vendors supplied both pieces to the puzzle. Now we're in a much different situation. Since the vast majority of network connectivity is via Ethernet, with some Token-Ring installations, vendors of emulation software no longer provide the hardware. With the advent of open-source software, we have different marketing models from which to choose. I divide them into three models: 1) closed source that has a price tag; 2) closed-source that is free for restricted use; and 3) open-source that is free for all uses. Let's examine each in more detail.

Model 1 software has been with us since the days of kerosene computers. This is the software with which we're all familiar and in which category we would place Client Access Express. Typically, solutions that fall into this category are full-featured and, in many ways, aren't much different from the word processors I mentioned earlier (that is, they have tons of features that, if needed, are a godsend and worth every penny you pay for the software). Most of them provide much more than simple terminal emulation, and some of them will run even on Unix machines. MCPressOnline's Vendor Directory lists over two dozen products in its "Emulation - Software" database. So there's no dearth of products available, in spite of the free alternatives.

Model 2 software is a bit more creative in its marketing plan. One example pointed out to me is available from Ericom. Ericom's software is closed source, but it runs on Linux. The hook? If all you want is simple 5250 emulation, the product is free. If you want printer emulation, you need to pay for it. The beauty of this model is that you pay only for features in excess of the basic emulator. And commercial support is available, although you'd need to actually purchase the product to get tech support. Since commercial support often is listed as one of the reasons to avoid open source, this should make the IT directors happy.

Finally, there's marketing model 3 software, which, obviously, is where I concentrate my attention. As I mentioned last month, there's the tn5250 project, the focus of which is to develop a fine emulator that runs on any *nix look-alike as well as Windows. The developers have also provided printer emulation along with secure communications, so it has all the basics. If you're looking for hot-spots and sophisticated cut-and-paste, you might want to look at the Java-based open-source project tn5250j. It does the basic emulation, but since tn5250j is written in Java, it is by definition cross-platform. Hence, it will work on any platform for which a Java Virtual Machine (JVM) is available.

I think the quintessential difference between software products that fall into the first two models versus the third open-source model can be found on the home pages of their respective products. Specifically, on the tn5250 and tn5250j sites you'll find links to each other. I doubt you'll find the same cooperation between the commercial ventures! What's important is that open source is about choice and flexibility. No commercial ventures were killed in the production of open-source software. And, perhaps, the very existence of open-source has improved the products available from everyone.

Next month, I'll continue my connectivity articles with Linux, ODBC, and you. Until then, have a great month. And keep those emails coming!

Barry Kline is a Linux and open-source evangelist. He can be reached by email at This email address is being protected from spambots. You need JavaScript enabled to view it..

Barry L. Kline is a consultant and has been developing software on various DEC and IBM midrange platforms since the early 1980s. Barry discovered Linux back in the days when it was necessary to download diskette images and source code from the Internet. Since then, he has installed Linux on hundreds of machines, where it functions as servers and workstations in iSeries and Windows networks. He co-authored the book Understanding Web Hosting on Linux with Don Denoncourt. Barry can be reached at This email address is being protected from spambots. You need JavaScript enabled to view it.

Also Read

Practical OSS: MkDocs for the IBM i, Part 2

Practical OSS: MkDocs for the IBM i, Part 1

Practical OSS: MMAIL from EASY400