The Four Hundred

Feelings about the creation of an open source version of OS/400 run strong, at least if the letters we have received from readers of The Four Hundred are any measure. Many readers were either for or against the so-called Open Source/400. After reading my article "Open Source OS/400: A Crazy Idea for Crazy Times," some readers were hopeful that it might inject a little extra pizzazz into the platform. Here are some of those letters.

Reading all those articles about Linux and open source, I'm always intrigued by the different stories they tell us about how it got that far, and what caused their success; still, I haven't found my theory in any of those.

If we look at the open source projects and their implementation in commercial environments, I guess their scale is still much too small to justify any of the success that is currently attributed to Linux or open source (let alone compared with the amount that was invested in it, even if it is "free").

Further, if we look at open source projects related to kernel development, the scale is even smaller (maybe apart from driver support, but with restricted hardware like iSeries, we wouldn't benefit much from this), and I doubt that, with the complexity of OS/400, we would find many people that have the level (realizing that many are still struggling with RPG III) of such development.

So what makes Linux popular, and is open source really popular? In my opinion, the reason Linux is popular is that it is free, not that it is open source. Or do you think that all those school kids (of which 99 percent have never looked at the source) would pay a similar (license) price as Windows to install an open source operating system just to get Linux?

So, if we think further, we would ask if we need Open Source/400? No, in my opinion. Not at all. We just need a cheap (and that's 50 euros/dollars for the media, not the amount that IBM has in mind) version of OS/400 that runs on Intel iron. Just imagine all those school kids that could install OS/400 for free on their hardware they have at home (and that's not an additional PowerPC chip for $2,500 that IBM would like to sell).

It will probably remain my theory, but if we want acceptance and a huge user base, this is, in my opinion, the only way. Making OS/400 open source, or installing Linux on iSeries, or whatever-- it won't bring iSeries to the schools! It's still too expensive for that, and doesn't run isolated on a single Intel PC.

Good article. I am amazed by the lack of awareness of this great machine. I often hear college kids make comments like "Mainframes are dead." My reply is always, "Name another system, mainframe or not, that can support Windows 2000 (internally and externally), Linux, Unix, OS/400, and that has been 64-bit RISC since 1995." I love the blank stare I get, usually followed by "Huh, it can do that?"

Are you aware of any Web sites or news groups dedicated to starting the ball rolling on this?

Thanks for your article. Since it's not April Fools' Day, I don't have to worry about the "gotcha!"

It's hard to imagine anything that would be worse for AS/400-based systems and the businesses that use them than OpenSource/400, or, as one of my colleagues refers to it, "open sores." IBM would be better to just give it away and retire support for the platform.

Businesses that have a clear picture of their IT TCO and ROI should not have a problem with using the AS/400, despite some of the pricing and licensing schemes used by IBM. Improved marketing, competitive pricing, and a direct appeal to business management would probably be a better step for IBM.

Regardless, I am very impressed by your reasoning and intentions expressed in the article. From my perspective, issues regarding the activism of some open source advocates and the implications for business computing are complex, and have at their root overtones that go far beyond a discussion of systems technology.

Timothy, I've followed your articles for years now and would rate your publications as the best in the industry. Thanks again for your article, keep up the fine work, and all the best with your future endeavors.

Open sores is very funny. I will definitely use that one from now on. I have limited personal experience with open source Linux or Unix, but millions of people are compelled to use these products. I think open source is potentially a great way to get people interested in OS/400 for OS/400's sake. Young programmers and techies grow up to be tomorrow's MIS managers and systems analysts, and they are not getting any exposure to OS/400 in high school or college. I wonder just how many of the millions of Linux licenses that are distributed (either sold as a compiled program or downloaded for free) end up behind the keyboards of those young techies. I think my life would have been a whole lot different if I had something like Linux at my disposal when I was 15.

You make your habits early in life, and the limited capabilities and high cost of early PCs kept me from getting as interested in programming as I might have been. And I had access to more computing power than most kids of my generation. When I was a teenager, I had a home-built Z80 machine from Ohio Scientific, with 8-inch floppy drives, that was the size of two milk crates. My best friend had one of the first Apple machines. His uncle, who was a System/38 programmer and then an AS/400 programmer, bought both machines because he had nothing better to do with his money than blow it on some nerdy teenagers. When my friend got the Apple, I got the milk crates. I learned a lot, but you couldn't really do much on a Z80. I wanted to run speech-recognition software and create simulations, and these machines just didn't have enough iron. So my interest drifted in other directions.

Now think, as many of you have, about how different the OS/400 market would be if OS/400 were as ubiquitous and cheap as Linux, and if it ran on just about anything? To be sure, IBM would sell the flagship iSeries iron to run the production OS/400 applications, but many of us might have cut our teeth on much smaller OS/400 machines long before we got to a position of buying or recommending an OS/400 platform for real work.

Thank God for IBM. Otherwise, we might all be running Unix. Why not list the bad reasons for open source?

Good idea. I think I will. When one vendor controls an operating system from top to bottom, that vendor can determine what does, and does not, go into that platform, and it can, if necessary, change direction and add features and functions in a very rapid manner, because it is not relying on the kindness of strangers, who work in their free time, to get code built for those new functions. These vendors have absolute control, and that makes them predictable and makes support more manageable. Also, most providers of major operating system platforms have decades of experience in coding operating systems with very strict quality standards that are measurable; these vendors often also require, as IBM does with OS/400, that the quality of each release should improve upon the prior one. There is also a consistent team of programmers, who often grow alongside the operating system as it develops over the years. There is no guarantee that open source OS/400 will, like Linux, spawn a tightly knit community of developers who are dedicated to nurturing and protecting OS/400.

Interesting article. I have several concerns about OS/400 becoming OpenSource400, two of which are business stability and the ability of IBM to support the operating system. Given a choice, I do not want OS/400 open source.

On the other hand, do I want OS/400 running on PCs? Absolutely. Will IBM do this? Unlikely.

The idea of Open Source/400 is excellent. This is the only chance to promote the life of OS/400. It looks like IBM lost hope on OS/400 and IBM itself is killing it. Thanks.

My one big objection to any open-source operating system is the support nightmare that can result from having fifty thousand different variations of an operating system in existence at the same time. I am not even convinced that Linux is viable yet; I think it will need some kind of a worldwide standards committee before people jump on the bandwagon wholesale. I think that the only reason Linux is being used today is because the various flavors are under the umbrellas of companies like IBM that can offer a version of it that is pretty much guaranteed to run. If there were no companies providing a "standard" version of it, it would remain the plaything of the live-eat-breath-sleep techies, who see technology not in terms of return on investment, but rather in terms of return on ego.

One of the great features about the iSeries is the reliability of its proprietary operating system. In the eight years that I have worked at my current position, we have not had either of our AS/400s crash because of an operating system problem (or any other software problem, for that matter). This is why all of our mission-critical applications are on AS/400s instead of Windows servers (and why Microsoft for many years ran its core applications on AS400s). If the source code is opened up to programmers/hackers/twiddlers to tweak at will, then IBM will have to either wash its hands of supporting OS/400 or start charging its customers by the hour (a more likely scenario) to figure out why their particular version of the system crashed or why it will not IPL. In either event, under an open model, we would have to start allocating expensive resources to keep the operating system functioning properly, especially if we started pulling in patches and fixes from the open-source community to incorporate into our installation (not to mention what would happen if we pulled in a virus). This does not even take into account the cost to our company if our system went down because of a glitch in our operating system. By contrast, under the current closed model, one of our programmers devotes a small percentage of his time to keeping the PTFs up to date, and we give little thought to the operating system, instead concentrating on application software development and enjoying the benefit of a system that never crashes. Because we do view technology in terms of return on investment, I vote in favor of keeping OS/400 closed.

There you go again, thinking outside the box! I have only one disagreement with your reasons in favor of open source OS/400. The legions of AS/400 programmers, I suspect, don't include many who understand C or Java. For that reason, and because we are very busy, I don't see many of us jumping into the guts of OS/400 to improve something. Also, a lot of us start to glaze over when we get very far into the operating system, since our stock in trade is applications.

My other concern is with how distribution and support would work (I haven't worked with Linux). How would the "official" version of OS/400 come together, unless IBM said, "Send us your codelets and we will integrate and test them." That would lead quickly to making choices about which changes are technically worthy and support a general need. Which could even lead to bureaucracy if IBM became the gatekeeper.

It would not be realistic to ask the typical AS/400 shop to evaluate and integrate a lot of codelets of unknown quality. They don't have the time and are averse to risking system stability.

IBM will definitely have some concerns about opening the kimono on anything that would make hacking easier.

Even with these (and other) concerns, your vision of a community of free black-belt coders enhancing OS/400 is enticing. They would first have to catch the vision, if you are talking about college students, and regard the OS/400 as open enough to be interesting. The commercial success of Unix, in spite of its awkward, arcane commands, is evidence of the power of enlisting students to understand your system. I once used Unix in a university course on C programming and thought it would never appeal to anyone who shaves regularly, but when the beard came off the techie knowledge persisted!

Exactly! Open Source/400 is not about letting everyone in the world create OS/400, but letting this generation of elite programmers participate today, and cultivating the next generation of newbies to think of OS/400 as, well, a normal thing to be playing around with, instead of Nintendo or Playstation consoles. I want OS/400 to be normal, familiar, inexpensive, and, therefore, a volume play with an easy sell. I do not want it to be an exotic technology with a complicated TCO-ROI play, with the consequent low volumes. Moreover, if there are things that the OS/400 community wants in OS/400, an open source approach means that they can code it themselves, if they really want to, and not sit around and wait for IBM to get the job done. OS/400 is mostly C and Java, after all, and there are no shortages of programmers in these languages. Creating the culture and systems to manage such a complex, worldwide, open development project would be troublesome, no question about it. But the technology--software change management tools, collaboration tools, standards bodies, and the Internet--exists to make it happen.

My thanks to these readers and to the many others who took the time to send in their thoughts on the possibilities of Open Source/400.

Sponsored By

BCD INT'L

Develop new or extend your existing iSeries-AS/400, BPCS, MAPICS, JDE or other applications, quickly and affordably, into elegant Web/Browser based applications.

Unlike many IBM AS/400 WebSphere™ sites, 95+% of BCD WebSmart users DO NOT have to upgrade their hardware, saving them $20K to $200K in costs!

====== Save $1,000 or more now. ======

Try Award Winning WebSmart, the complete development and deployment environment tool, on a FREE DOWNLOAD. Prove it out quickly. Prove the quick learning curve; prove the speed of the developed applications on small, medium or large sized AS/400s. Develop applications up front, risk free, and without obligation, then decide. Buy with confidence.

WebSmart developed ILE/CGI web/browser based applications DO NOT consume interactive resources and run significantly faster than applications generated in JAVA--significantly faster. Runs really fast even on small iSeries-AS/400s. NO Proprietary data access middleware, no hidden additional costs and unlike with NT servers--NO viruses.

You will soon see why BCD won 18 industry awards, including 10 awards for best Web, and iSeries-AS/400 application development tools. BCD has 1,000's of developers and 500,000+ end-users using the developed apps.

http://www.bcdsoftware.com

====== Read what the EXPERTS have to say ======

-- Regarding the WebSmart developed application Formica created: "We are rolling out to new distributors every day, and the feedback has been very positive . . . I can safely say this project was a success. . . "We're serving about 16,000 unique page hits and 18 MB of data everyday."
Phil Catlin, Application Development Manager, Formica Corp.

-- "It took only four days from not having ever seen the software to a complete production application that was ready to be used by agents and customers."
Lee Gillow, Airways Freight

-- "With the reasonable acquisition cost of WebSmart it was easily cost justified. Our customers love it and we are very pleased with the end result."
Dolly Newville, MIS Manager, Weigh-Tronix

WebSmart users include: Altec Lansing, Affinity Insurance, Arrow Environmental, Behr Process, Baker Petrolite, Bristol Hospital, Calvin Klein, City of Moorhead, Diesel, D.J. Powers, Dole-Standard Fruit, EMED, GasBoy Int'l, General Semiconductor, Goodyear Tire & Rubber, Hoshizaki America, Hytrol Conveyer, Lamps Plus, Kauai Electric, Kelly's Kids, Kwik Trip, Labette Community College, Legacy Partners, Midwest Trophy, MSU, Nacco Materials, Oregon Dept. of Vet. Affairs, Peg-Perego, Penn Fishing Tackle, SouthTrust Bank, State of California, Testor Corp., TST/Impresso. . .

====== WebSmart Overview ======

WebSmart offers you Superior Technology that lets you create your own solutions, retain development control, plus have the FREEDOM of CHOICE.

- Create, extend and deploy fast, fully secured, elegant B2B, B2C Web/browser based apps.

- Run on either the iSeries 400 HTTP or Apache HTTP server (ILE/CGI), and/or on multiple platforms such as Windows (NT/2000), Linux, UNIX using WebSmart's JSE (Java Servlet Edition) that runs on WebSphere, Apache Tomcat, BEA Weblogic. . . servers. All from ONE SINGLE IDE EFFORT!

Develop new or customize the 70+ FREE templates and e-Commerce apps with the PC based IDE. Extend any iSeries-AS/400 app. Price includes unlimited developer and end-user seats. A developer can have a seat at work, on their laptop traveling, and at home. Work anywhere on or off-line (with syntax checking, validation. . .), email objects for true portability. Use existing skills, requires little or no Web or iSeries-AS/400 programming, and. . . it's affordably priced.

====== FREE Download & Special Promotion ======

LIMITED TIME PROMOTION: Purchase WebSmart and get 10% off the list price, plus 30% off any other BCD product that is ordered within 45 days of a WebSmart purchase, including these: ProGen Plus, Nexus, Catapult, Spool-Explorer, Docu-Mint. . . (20% off on EZ-Pickin's). No prior sales.

For a complete evaluation get the FREE DOWNLOAD and professionally written user guide. Click http://www.bcdsoftware.com, email sales@bcdsoftware.com or call us at 630-986-0800 to discuss your needs and how WebSmart can help solve them. All BCD products come with excellent support.

BCD, for over 25 years, a name you can trust: Winner of 18 Industry Awards, 9,500+ customers and 23,000+ products sold.

==================================================

All trademarks mentioned are trademarks of their respective owners.